curriculum vitae pulakesh chandra maiti professor of statistics, …pulakesh/cvpmaiti.pdf ·...

80
Curriculum Vitae Pulakesh Chandra Maiti Professor of Statistics, Indian Statistical Institute E-mail: [email protected] , [email protected] 1. Personal Details: 1. Name : Pulakesh Chandra Maiti 2. Date of birth : 21 st October, 1950 3. Citizenship : India 4. Present position/designation: Professor, Indian Statistical Institute 5. Address: (a) Office 203, B.T. Road, Kolkata – 700108 Phone: 91 33 25752612 E-mail: [email protected] , [email protected] (b) Residence: Krishna Apartment Flat – 3A 8/1B, Panditia Road Kolkata – 700029 Phone: 91 33 2476 4170 (Land); +919432304116 (Mobile) 2. Academic and Professional Details: 1. Qualifications: Degree/Diploma Subject Institution Year Statistician’s Diploma Statistics Indian Statistical Institute 1974 Ph.D Sampling Indian Statistical Institute 1983 3. Permanent Positions: (a) Technical Assistant, Sankhya, Indian Journal of Statistics, Indian Statistical Institute, July 78 – April 1983. (b) Consultant, Market Research Survey, Operations Research Group (ORG), Vadodara, April 14, 1983 – October 23, 1984. 1

Upload: others

Post on 26-May-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Curriculum Vitae Pulakesh Chandra Maiti

Professor of Statistics, Indian Statistical Institute E-mail: [email protected], [email protected]

1. Personal Details: 1. Name : Pulakesh Chandra Maiti 2. Date of birth : 21st October, 1950 3. Citizenship : India 4. Present position/designation: Professor, Indian Statistical Institute 5. Address: (a) Office 203, B.T. Road, Kolkata – 700108 Phone: 91 33 25752612 E-mail: [email protected], [email protected] (b) Residence: Krishna Apartment Flat – 3A 8/1B, Panditia Road Kolkata – 700029 Phone: 91 33 2476 4170 (Land); +919432304116 (Mobile) 2. Academic and Professional Details:

1. Qualifications:

Degree/Diploma Subject Institution Year Statistician’s Diploma

Statistics Indian Statistical Institute 1974

Ph.D Sampling Indian Statistical Institute 1983 3. Permanent Positions:

(a) Technical Assistant, Sankhya, Indian Journal of Statistics, Indian Statistical Institute, July 78 – April 1983.

(b) Consultant, Market Research Survey, Operations Research Group (ORG), Vadodara, April 14, 1983 – October 23, 1984.

1

(c) Publication Officer, Sankhya, Indian Journal of Statistics, Indian Statistical Institute, October 23, 1984 – November 30, 1990.

(d) Lecturer, Economic Research Unit, Indian Statistical Institute, November 30, 1990 – June 1, 1995.

(e) Associate Professor, Economic Research Unit, Indian Statistical Institute, June 1, 1995 – September 1, 2014.

(f) Professor, Economic Research Unit, Indian Statistical Institute, September 1, 2014 till date.

4. Visiting Positions:

(a) Visiting Scientist (August 3 – August 15, 2014) The Institute for Mathematical Research (INSPEM) University Putra, Malaysia.

(b) Visiting Professor (April 10, 2013 - May 02, 2013) Department of Statistics University of Rajshahi, Bangladesh

(c) Visiting Professor (August 28, 2006 – September 8, 2006) UNIVERSIDADE NOVA DE LISBOA Lisbon, Portugal

(d) Senior Consultant (May 22, 2000 – May 21, 2001) National Council of Applied Economics Research New Delhi.

(e) Visiting Professor (1996 – 1997) PG Department of Statistics The MEDUNSA Pretoria, South Africa

(f) Chief Research Specialist (Sampling): (January 27, 1995 – December 15, 1996) Centre for Statistics Human Sciences Research Council (HSRC) 134, Pretorious Street Private Bag X41 Pretoria, South Africa

(g) Guest Lecturer (April 1, 1993 – March 1995) University of Calcutta, Kolkata.

2

5. Honorary Positions:

(a) In Charge of One Year Evening Course in Statistical Methods and Applications (1990 – 93) Indian Statistical Institute held at different Centres, Kolkata, Delhi, Bangalore in India.

(b) Editorial Secretary, Sankhya Indian Journal of Statistics, Indian Statistical Institute, Kolkata (1991 – 1993).

(c) Member, Working Group 69th Round NSS (July 2012 – December, 2012), National Sample Survey, Appointed by National Statistical Commission, Government of India.

(d) Member, Working Group, 73rd Round NSS (June 2015 – June 2016 ) National Sample Survey, Appointed by National Statistical Commission, Government of India.

(e) Member, Working Group, Drug Abuse Survey (November 18, 2014 – April 10, 2015), Ministry of Statistics & Programme Implementation, National Statistical Commission, Government India.

(f) Member, Expert Group, 11th Five Year Plan, (April 2007 – March 2012), Preparation of Annual Perspective Plan (Under the Process of Decentralised Planning) District of Howrah, West Bengal, India.

(g) Analyst of village resource mapping. (2014) Directorate of Micro and Small Scale Industries, Government of West Bengal, India.

(h) Member, “Technical Assistance Piloting below poverty line”, (May 2010 – 2011): (Census) Panchayat & Rural Development, Government of West Bengal, West Bengal, India.

(i) Collaborative Scientist (May 17 – August 16, 2010), Department of Economics and Statistics (DES), Tata Service Limited, Mumbai, India.

(j) Member, (2006) Uganda Bureau of Statistics, Uganda. 6.1. Services to Government of India:

• Serving as a sampling expert on the working group of 69th round of NSS, which was devoted to Drinking Water, Sanitation, Hygine and Housing conditions (including slums), sampling design, methodology and strategies to reduce non-sampling errors were developed. Sampling design, schedules of enquire and procedures for data collection were finalized at all India Training of Trainers (AITOT) New Delhi, during March 15 – 16, 2012.

3

• A special estimation procedure of estimating the number of slums in the absence of a list of slum was developed. The sampling methodology thus developed suggests the number of slums can be estimated unbiasedly from the survey with the help of an additional information on the number of blocks linked to a slum of which the sample blocks are found to be parts and the methodology thus developed was deployed in the 69th round. Number of slums at the state as well as at all India level was estimated.

• On the pilot basis, field visits were made during /and after the NSS survey at the states namely Tripura, West Bengal, Jammu and Kashmir, Gujarat, Andaman & Nicobar islands to make an assessment of related field problems including problems of non-sampling errors in connection with 69th round NSS. The report was submitted to the office of National Statistical Commission, New Delhi.

• Served as a Chairman of the subcommittee of 69th round on imputation of rental values, and the report was submitted to the Commission.

• Provided knowhow on methodology, sampling design, and development of enterprise schedule in conducting nationwise surveys of NSS 73rd round devoted to surveying un-incorporated non-agricultural enterprises in manufacture, trade and other services (excluding construction). For the first time, Enumeration Block(EB) in urban areas was suggested as the first state unit (FSU).

• Provided knowhow on Drug abuse survey as a member of the working group. A dual approach i.e., household based survey for commonly used (licit) drugs and Respondent Drives Sampling Design (RDS) technique for illicit drugs was recommended in conducting drug abuse survey.

6.2. Services to Government of West Bengal:

• Created Data Base and for Creating the RDBMS, such studies as System Requirement Study (SRS), Context Analysis and Design (CAD), Data Flow Design (DFD) were made.

• Provided know how on Design and Development of Application Software programmes for processing and generating statistical distribution of the concerned variables. This lead to development of 105 software programmes.

• Development 157 booklets of Resource Mapping for 157 Gram Panchayats of the District Howrah, West Bengal.

• Prepared fourteen Human Development Reports at the block level of the fourteen blocks entitled “Block wise Human Development Report in the District of Howrah West Bengal.

4

• Provided guidance to the Directorate of Micro and Small Scale Enterprises regarding the fields to be considered for identifying potential resources in specific areas of different districts with the help of village resource information as obtained from Different District Industries Centre (DIC). This may ultimately facilitate industrilisation in the state.

• Provided know how on imparting training and supervision for the field work in conducting nationwide surveys, data analysis and inference in connection with the project entitled “Assistance of Piloting Below Poverty Line (BPL)”.

• Provided know how on projection of Demand of thirty two items of consumption under the project of the Department of Economics and Statistics (DES), Tata Services Limited, Mumbai.

• Provided expertise on Stochastic Modeling of Buying Bachviour of Indian Customers to Business Research and Corporate Planning Group of Hindustan Lever Limited and know how on Stochastic modeling of discovery of Hydro Carbon as needed by Oil & Natural Gas Commission, Government of India.

• Provided know how as a member in the consulting team for price-water house project “Strengthening Local Government in MP”. The project was highly appreciated so much so that it was reported in the ‘News Letter’ Section of the American Statistician.

• Provided expertise on statistical assistance to the Uganda Bureau of Statistics with respect to possible methodologies that could be implemented by the Uganda Bureau of Statistics, Uganda in such identified areas as (i) the estimation of yield rates of crops such as rice, potatoes and various fruits and (ii) estimation of total production in enterprises owned and operated by Government of West Bengal, (iii) estimation of consumption and expenditure on household items using Sample Survey Methodology.

6.3. Services to the Industry Houses: • Provided knowhow on methodology and applications of Statistics to Hindustan

Lever Limited, Tata Group of Consultancy Services in conducting nationwide surveys, data analysis and inference.

The project through which the above are realized are mentioned in detail under 4 of section 18.

7. Awards and Appreciations: 1. Received the “Leading Scientists Award, 2013” from International Biographic Centre, Cambridge, UK. 2. As a mark of appreciation, works on the projects entitled “Strengthening Local Government in Madhya Pradesh India (1998 – 1999), “Statistical

5

Information System for local planning by local bodies with creation of computerized data base for Howarh, West Bengal, India (April 2007 – March 2012)” and “Development of Consensus Development plan for the Gram-Panchayat Amardah under the Block Shyampur – II of the District of Howrah of West Bengal” have been kept in the Prasanta Chandra Mahalanobis (PCM) memorial Archive and Museum for preserving the historical heritage of the Institute for posterity. 3. The first ever excise in the country on decentralized planning at the Gram Panchayat Level entitled “Statistical Information System for local level planning” was carried out.

“Development of Concensus Development plan for the Gram Panchayat Amardah under the block Shyampur – II of the District of Howrah, West Bengal, India” has been preserved at the State Planning Board, Government of West Bengal, India.

4. Books on Human Development Report have been preserved in the National Library Kolkata under rule 3 of the delivery books (Public Libraries) Act, 1954 (27th May , 1954).

8. Invited Talks:

1. At the Tata Consultancy Service, Tata Group of Industries, Mumbai, (March 18, 2015).

2. At the International Conference on Recent Advances in Mathematics, Statistics and Computer Science (ICRAMSCS – 2015), Central University of South Bihar, (May 29 – 31, 2015).

3. At the International Conference on ‘Celebrating Statistical Innovation and impact in a world of Big and Small data, Department of Statistics, Sabitribai Phula, Pune University and International Indian Statistical Association, (December 20 – 24, 2015).

4. At the International Conference on Statistics and related areas for Equity, Sustainability and Development (SRAESD – 2015) & XXXV Annual Convention of Indian Society for Probability and Statistics (ISPS), University of Lucknow, (November 28 – 30, 2015).

5. At the National Seminar on ‘Poverty, Inequality and Health in India with Special reference to North East India’, North Eastern Hill University, and Indian Statistical Institute, (October 8 – 10, 2015).

6

6. At the 9th International Triennial Calcutta Symposium on Probability and Statistics, Department of Statistics, University of Calcutta and Calcutta Statistical Association, (December 28 – 30, 2015).

7. At the National Conference on Agricultural and Rural Development issues in Eastern India, Indian Statistical Institute, Giridih, (March 12 – 13, 2015).

8. At the 68th Annual Conference of Indian Council of Agricultural Research (ICAR), New Delhi (February 23 – 25, 2015).

9. At the National Conference “National Conference on Recent Trends and Development in Statistics, Department of Statistics, NCRTDS”, M.D. University, Hariyana, (February 21 – 23, 2015).

10. At the 17th Annual Conference of the Society of Statistics, Computer and Applications (SSCA) on ‘Statistics and Informatics for Smart decisions in Management Resources: issues and challenges, Birla Institute of Management Technology, (February 23 – 25, 2015).

11. At the International Conference on Recent Advances in Mathematical Statistics and its application in Applied Sciences, Department of Statistics, Gauhati University, (December 31 – January 1 -2, 2014).

12. At the National Conference on Recent Advances in Statistical and Mathematical Sciences and their Applications (RASMSA – 2014), Kumayun University, (October 14, 2014).

13. At the 12th Conference of Indian Association of Social Sciences and Health (IASSH), GL Gupta Institute of Public Health, University of Lucknow, (November 21 – 23, 2014).

14. At the 2nd ISM International Statistical Conference 2014, with Applications in Sciences and Engineering (ISM – II, 2014), Pahang, Malaysia, (August 12 – 14, 2014).

15. At the Department of Statistics, Lucknow University, India (November 16 – 17, 2014).

16. At the International Conference on Recent Advances in Mathematical Statistics and its applications in Applied Sciences, Dhaka University, Bangladesh, (December 31, 2012 & January 1 – 2, 2013).

17. At the 1st International Conference on Information, Operations Management and Statistics (ICIOMS 2013), Kuala Lumpur, Malaysia, (September 1 – 3, 2013).

18. At the International Conference on Statistical Data Mining for Bio-informatics, Health, Agriculture and Environment, Rajshahi University, Bangladesh, (December 22 – 24, 2012).

19. At the 99th Session of Indian Science Congress Association, KIIT, Bhubaneswar, India(January 3 – 7, 2012).

7

20. At the International Workshop on Sample Surveys with special emphasis

on Large Scale National Sample Surveys towards promoting national development, Department of Statistics, Rajshahi University, Bangladesh (October 18 – 19, 2012).

21. At the Department of Statistics, Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur (January 10 – January 12, 2008).

22. At the conference of Central and State Statistical Organisations (COCSSO), Delhi (April 10 – 11, 2007).

23. At the Universidade Nova De Lisboa, Lisbon, Portugal (August 29, 2006 to September 01, 2006).

24. At the SCRA 2006 Conference, Lisbon, Tomar, Portugal (September 1, 2006 – September 4, 2006).

25. At the workshop on Demography and Population Dynamics with emphasis on Survey Sampling Methodologies, Tejpur University, Tejpur (October 26 – 28, 2006).

26. At the International Triennial Calcutta Symposium on Probability and Statistics, Department of Statistics, University of Calcutta and Calcutta Statistical Association (December 29 – 31, 2006).

27. At the national conference on ‘Recent Trends in Estimation and Optimizations”. Also to celebrate fifty years of Professor V.P. Godambe’s revolutionary theorem on unified theory of sampling, Institute of Science, Nagpur; Golden Jubilee (January 1 -2, 2006.

28. At the UGC sponsored refresher course on ‘Decentralisation, Planning and Participatory Development, Department of Economics with Rural Development, Didyasagar University, (December 12 – 31, 2005).

29. At the Training Programme Organised by Operations Research Guide India, ORG, India CA 242, Salt-Late, (December – 17, 2005).

30. At the Department of Economics, Jadavpur University, (December 31 – January 3, 2005).

31. At the national seminar ‘Information Support for rural development, Indian Association of Special Libraries & Information Centres (IASLIC), Golden Jubilee, (September 2004 – August 2005).

32. At the Tenth Course/Workshop on Sampling Design and Establishment Surveys, jointly with CSO, Ministry of Statistics and Programme Implementation and UN Statistical Institute for Asia and the Pacific, Japan, (October 18 – November 12, 2004).

33. At the joint IMS-SRMS Mini Meeting on current topics in Survey Sampling and Official Statistics, Calcutta, India, (January 2 – 3, 2004).

34. At the 275th IASLIC Study Circle meeting, (June 17, 2004).

8

35. At the eleventh International conference – SCRA 2004 – FMXI, Forum for Interdisciplinary Mathematics, Institute of Engineering and Technology, Lucknow, India (December 27 – 29, 2004).

36. At the UGC-sponsored national seminar on Participatory Decentralised Planning: Issues of Functions and Functionaries, Economics with rural development, Vidyasagar University, India (2003).

37. At the International conference ‘Recent Statistical Techniques in Life Testing, Reliability, Sampling Theory and Quality Control, Department of Statistics, Faculty of Science, Banaras Hindu University, India (December 29 – 24, 2003).

38. At the two day UGC sponsored national seminar on Decentralised planning, Department of Economics with rural development, Vidyasagar University, India (March 20 – 21, 2003).

39. At the Business Research Group and Corporate planning, Hindustan Lever Limited, Mumbai, India (August 12 – 13, 2003).

40. Keynote address speaker at the two-day national seminar on information support for rural development, conference hall of the Central Library, Vidyasagar University Campus, India (November 18 – 19, 2003).

41. At the seminar talk on Decentralised planning, Department Research Support Scheme (DRS), Vidyasagar University, India (November 07,2002).

42. At the XXI Indian Social Science Congress, Tamil University, Thanjavour, India (December 06, 1997).

43. At the Programme of Seminars, Department of Statistics, University of South Africa, First Semester, 1995, Pretoria, South Africa.

44. At the University of Transkei, Umtata, Transkei, South Africa, (March 28 – March 29, 1995).

9. Serving as a Resource Person: a. On a three day University Grants Commission (UGC) sponsored short

time course on “How to Write Research Project”, for the college and university teachers organized by Academic Staff College, Sambalpur University, Orissa, India, (January 22 – 24, 2013).

b. On a UGC-sponsored workshop in Mathematics, Nabadwip Vidyasagar College, Nadia, West Bengal, India (January 10, 2012).

c. On a UGC-sponsored workshop on Statistical methods, University College of Raiganj, Dinajpur, West Bengal, India (February 10 – 17, 2011).

d. On a UGC-sponsored refresher course (Statistics), UGC Academic Staff College, Sambalpur University, Orissa, India (December 4 – 24, 2008).

9

e. On a workshop on Large Scale Sample Survey and Regression Technique, Department of Statistics, (March 12 – 14, 2014).

f. On a weeklong workshop at the Institute of Regional Planning, Bhubaneswar, (June 9 – 14, 2014).

g. On a workshop on “Survey Methodology”, “Basic Principles of Survey Sampling”, Centre for Statistics, HSRC, Pretoria, South Africa (March 23 – 29, 1996).

h. On a workshop on Advanced Sampling Techniques, Centre for Statistics, HSRC, Pretoria, South Africa (August 14 – 18, 1995).

i. On a workshop on Statistical Distribution, Department of Economics and Statistics, Tata Services Limited, Bombay House, Mumbai (May 16 – 18, 2005).

j. On a workshop on how to use data collected through the Base line Survey Programme of the District of Howrah, West Bengal, India, (July 27, 2010 – October 13, 2010).

10. Dissertation Supervision: Supervised Masters dissertation in different curricula under different Universities.

11. Services to the Universities: a. Rasthrapati Tukadori Maharaj Nagpur University:

• Appointed as the External examiner for the Practical Examination in Pr-I (Statistics) at the MA/M Sc. II examination, 2006, 2007.

b. University of Burdwan:

• Appointed as a Post-Publication Review Examiner in paper VIII/I (Stat.& Eco.) for reassessing the scripts of MA/M.Sc. Economics, Part II examination, 2003, 2004, 2005, 2006 and 2007.

• Appointed as a Review Examiner in part II examination, paper VIII/1 (Stat. & Econ.) for MA/M.Sc. examination, 2003, 2005, 2006.

c. Vidyasagar University:

• Appointed as a moderator in Economics, MA/M.Sc. Part II examination 1998, 2000, 2001,2004, 2005, 2006.

• Appointed as a re-examiner of Economics, paper VIII for the M.Sc. examination 2003.

• Appointed as a paper setter for the Special paper XV (New) for the MA/M.Sc. part II examination 2004.

• Appointed as a re-examiner on economics paper VI for the MA/M.Sc examination, 1998, 1999, 2001, 2002.

10

12. Academic Administration: 1. Member convener of the organizing committee for International

Conference on Environmental and Ecological Statistics with Applications as a part of Platinum Jubilee of the Indian Statistical Institute, March 21 – 23, 2007.

2. Member, JRF Selection Committee for the Academic Year 2003 2004. 3. Expert member of the Selection Committee to the post of Assistant

Professor for Visva-Bharati University, Central University, (2012), India. 4. Expert member on Research Methodologies for Mahatma Gandhi National

Rural Employment Guarantee Act (2013), Ministry of Rural Development, Government of India.

13. Member of Scientific Bodies: (a) Life member, Sankhya, Indian Journal of Statistics (L/5222); (b) Life member, Calcutta Statistical Association; (c) Life member, Asiatic Society (M00250); (d) Life member, Indian Association of Social Science and Health (IASSH); (e) Life member, Indian Science Congress Association (418904).

14. List of Publications: 14.1. Journal of Publications:

1. Interactive Linear Models in the Context of two Stage Sampling, To appear in Volume 14, No. 1, March 2015.

2. Interactive Linear Models in Survey Sampling (2013) (with B.K. Sinha), JSTA, Volume 13, No. 3. pp. 263 – 272.

3. Estimation of Nonsampling Variance Components Under the Linear Model Approach (2009). Statistics in Transition, New Series, Vol. 10, No. 2, pp 193 – 222.

4. Some improved variance estimators from a bivariate non-normal population (2009). (with T.P Tripathi). Pakistan Journal of Statistics, Vol XX(X) pp 165 – 194

5. A note on optimum inclusion probabilities in WOR-sampling scheme based on super population model (2009). (with T.P Tripathi), Pakistan Journal of Statistics, Vol XX(X), pp 221 – 226.

6. Existence of the BLUE for finite Population Mean under multiple imputation (2008) Statistics in Transition – new series, Journal of Polish Statistical Association, Vol 9, pp 233 – 258

7. Stochastic Modeling of Buying Behaviour of Indian customers (2008), Calcutta Statistical Association Bulletin, Vol 60, pp 111 – 112.

11

8. Some Shrinkage-Type Estimators for Variances of Univariate Populations (2007): To Appear in JISPS (Journal of Indian Society of Probability and Statistics) in 2007 issue.

9. Unbiased Estimation of a Finite population: Case of Multiple Indirect Identifiability of Population Units (with B. K. Sinha and S. Sengupta) (2006): JSTA (Journal of Statistical Theory and application), Volume 5, 1, pp. 81-90.

10. Information Support for Planning and Rural Development (2004): IX, Journal of Economics, Volume IX, Vidyasagar University.

11. Statistical Enquiry in identifying some specific castes as OBC’S in West Bengal, India, (2000): Bharatiya Samajik Chintan, XXII, 1-2.

12. Stochastic Modeling and Forecasting of Discovery, Reserve and Production of hydrocarbon with an Application (with J.K. Ghosh, et. al.) (1997): Sankhyā, Series B, Volume 59, Part 3, pp. 288-312.

13. An Econometric Model of Exploration and Exploitation of Hydrocarbon (with M. Pal) (1997): Journal of Quantitative Economics, Volume 13, No. 2, pp. 29-44.

14. Evolution of Statistics in India (with J.K. Ghosh et. al.) (1997): International Statistical Review, Volume 67, 1, pp. 13-34.

15. On Minimax Allocation of Stratified Random Sampling when only the order of Stratum Variances is known (with M. Pal) (1994): Statistics and Decisions, vol 12, pp 195 – 201.

16. Estimating unknown dimensions of a binary matrix with applications to estimation of the size of a mobile population (with M. Pal and B. K. Sinha) (1992): Statistics and Probability, Edited by S. K Basu and B. K Sinha.

17. Some results on T1-class of linear estimators (1988): Jour. Indi. Society of Agri. Statistics, vol LX, No.1, pp 1 – 8.

18. Estimation of Lorenz Ratio from a finite population (with M. Pal) (1988): Anvesak, vol 18, Nos. 1 – 2, June – Dec, pp 29 - 62.

19. On some estimates of Poverty Measures (with M. Pal) (1988): Calcutta Statistical Association Bulletin, vol 37, March & June, Nos 145 – 146, pp 81-90.

20. A unified approach to estimation of Lorenz Ratio from a finite population (with M. Pal) (1988): Sankhya, Series B, vol 50, part 2, pp 215 – 223.

21. Use of prior information on some parameters in estimating population mean (with T.P Tripathi, S.D Sharma) (1983): Sankhya, Series A, vol 45, part 3, pp 372 – 376.

22. A note on the Estimation of Mean Square Error (1982): The Aligarh Journal of Statistics vol 2, pp 38 - 40.

23. Some T2-class of estimators better than H-T estimator (1981): The Aligarh Journal of Statistics vol 1, No.1, pp 52 - 58.

14.2. Papers under Communication: 1. Revised version of the paper entitled “The Indian Official Statistical System

Revisited” by P. Maiti., T.J. Rao and J.K. Ghosh: Submitted to Sankhya Series B.

12

2. Mean Square Error Decomposition Model under two-Stage Sampling: Estimation of Population Mean: by P. Maiti and Mahendra S. Shiton. Submitted to Statistics in Transition, New Series.

3. Estimation of Measurement Variance Under two stage sampling: Estimation of Population Mean by P. Maiti. Submitted to Statistics in Transition, New Series.

14.3. Published in Proceedings of Conference: 1. A Generalised Horvitz-Thompson Estimator for Population total with an

application in estimating number of species in the Region of Silent Valley in India (2015): National Conference on Recent Advances in Statistical and Mathematical Sciences and their Applications (RASMSA – 2014), Kumayan University, India, October 2014.

2. Estimation of Measurement Variance in the Context of Environment Statistics (2014), 2nd ISM International Statistical Conference 2014, with applications in Sciences and Engineering (ISM – II 2014), Pahang, Malaysia, August 12 – 14, 2014.

3. Estimation of variance of Horvitz-Thompson estimator in the presence of

measurement error: The linear model approach (2013): 1st International Conference on Information Management and Statistics (ICIOMS), September 1 – 3, 2013, Kuala Lumpur, Malaysia.

4. Some Aspects of Decentralised Planning International Conference on Recent

Advances in Mathematical Statistics and its Applications in Applied Sciences in Collaboration with Indian Statistical Institute, Kolkata; Department of Statistics, Gauhati University DST FIST and UGC(SAP) DRS-I, Department, Gauhati 781014, Assam, India, December 31, 2012 – January 1 -2, 2013.

5. Interactive linear model in survey sampling (2013): 7th Annual Conference on

Statistics, 17 – 20 June, 2013, Athens, Greece. 6. Estimation of finite population total: Interactive Model in Survey Sampling with

Bikash Sinha, International Conference on Statistical Data Mining for Bio-informatics, Health, Agriculture and Environment , December, 21 – 24 2012, Proceeding Department of Statistics, University of Rajshahi, Bangladesh.

7. Information created and developed for decentralized planning to two different blocks

of the district of Howrah, West Bengal (2012): International workshop on large scale national surveys pp. 20 – 28, Department of Statistics, University of Rajshahi, Bangladesh ISBN 978-984-33-7470-77.

8. Information for Decentralized Planning in the session of Mathematical Sciences: case

studies and surveys in the 99th Indian Science Congress during January 3 – January 7, 2012, held at KIIT University, Bhubaneswar, Orissa, India.

13

9. A Tale of Two Externally Funded Project Reports: M. P. Experience (Jointly with J. K. Ghosh and B. K. Sinha) (2007), Presented at the Conference of Central and State Statistical Organizations (COCSSO) held during April 10-11, 2007.

10. Some aspects of unbiased estimation of size of tree species in the Western Ghats of

Western India (2007): Presented at Platinum Jubilee Conference on Environment and Ecological Statistics, March 21 – 23, 2007, Indian Statistical Institute, Kolkata.

11. Stochastic Modeling of Life Data (2005): Presented as an invited lecture in

Department of Economics and Statistics, Tata Services Limited, Mumbai, May 16-18, 2005.

12. Non-Sampling Errors – Classification, Quantification and Decomposition of MSE of

a Survey Based Estimator (2003). Presented at 5th International Calcutta Symposium on Probability and Statistics, December 28-31, 2003.

13. Improved variance estimators from bivariate normal population (with Dr. T. P.

Tripathi) (2003): Presented at the International Conference on Recent Statistical Techniques in Life Testing, Reliability, Sampling Theory and Quality Control, Benaras Hindu University, December 29-31, 2003.

14. Combined use of Method of Moments and of Generalised Least Squares in

estimating the parameters under inverse sampling (1997): Presented at the Conference on Recent Advances in Statistics and Probability, Dec 29, 1997 – Jan 1, 1998, Indian Statistical Institute, Kolkata.

15. Sampling and Estimation Procedures for Inverse Multinomial sampling associated

with single, multiple and joint events (with H. S Styen) (1996): Presented at the Annual Conference of SASA, at the University of Stellenbosch, Nov. 6 to Nov. 8 1996.

16. Asymptotic Behaviour of two estimates of Poverty Measures (1995): Presented at

the Annual Conference of the South African Statistical Associations (SASA), Nov. 5 1995, University of Orange Free State (OFS).

17. The Literacy program in India and its evaluation (1995): Presented at the Department

of Statistics, Unversity of Transkei, March 29. 18. My experiences with Practical Sampling (1995): Presented at Department of

Statistics, University of Transkei, March 28, 1995 and University of South Africa (UNISA), Feb 21, 1995.

19. A Stochastic modeling on the discovery of Hydro-carbon with application to Indian

data (1993): Presented at the Annual Conference of Operation Research Society of India, Calcutta.

20. The process of Exploration and Exploitation of Hydro-carbon (1993): Presented at

P. C. Mahalanobis Birth Centenary Celebrations Symposium on Sample Surveys: Theory and Methods, Dec 15 – 17, Indian Statistical Institute, Kolkata.

14

21. An econometric approach to the estimation of hydro-carbon (1992): Presented at the

Annual Conference of Operation Research Society of India, Ahmedabad, India. 22. Some results on T1-class of estimators (1980): Presented at the Conference of

Mathematical Statistics and Probability Theory in Honour of Prof. C. R. Rao to mark his 60th Birthday, New Delhi, India.

23. Use of prior information on some parameters in estimating population mean (with T.

P Tripathi) (1978): Presented before 33rd Annual Conference of the Indian Society of Agricultural Statistics, Trichur, India.

24. The use of multivariate auxiliary information in the selecting the sampling units (with

T. P Tripathi) (1976): Proceedings of the Symposium on Recent Developments in Survey Methodology. Indian Statistical Institute, Kolkata.

14.4. Books Published:

a) Some Aspects of Complex Design in Survey Sampling (1996):

HSRC/RGN Publishers; ISBN 0-7969-1857. A brief summary has been provided.

b) Sampling and Estimation Procedures for Inverse Multinomial Sampling Associated with Single, Multiple and Joint Events (with H.S. Steyn) (1997): HSRC/RGN publisher, 134, Pretorious Street, Pretoria 0002, ISBN 0-7969-1822-8. A brief summary has been provided.

c) Sampling and Estimation Procedures for Inverse Multinomial Sampling Associated with Single, Multiple and Joint Events by H. S. Steyn and P. Maiti, HSRC / RGN Publishers; ISBN 0 – 7969 – 1822 – 8.

d) Basic Principles of Survey Sampling (1996): CENSTAT, HSRC/ RGN, Pretoria, South Africa.

e) Survey Methodology (1996): CENSTAT, HSRC/RGN, Pretoria, South Africa.

f) A Unified Set up for Probability Sampling (1996): CENSTAT, HSRC/RGN, Pretoria, South Africa.

g) Statistical Information System for Decentralized Planning by Local Bodies in the District of Howrah (2008). Howrah Zilla Parishad, Govt. of West Bengal.

15

15.1. Summaries of the Research Papers (Published in referred journals): (1981-1994 &1995 onwards):

1. Interactive Linear Models in the context of two-stage sampling. To appear in Volume 14, No. March 2015. Summary: Following Sinha and Maiti (2014), we continue our investigations along a linear interactive model by incorporating a two-stage sampling design. Expressions for an unbiased estimator of the finite population total along with its unbiased variance estimator have been derived. Essentially, it incorporates second level of randomness.

2. Interactive Linear Models in Survey Sampling (2013) (with B.K. Sinha), JSTA, Volume 13, No. 3, pp. 263 – 272. Summary: Considered is a liner interactive model in the context of survey sampling. The situation arises when investigator and/or supervisor interventions are contemplated in the responses. An appropriate linear model is introduced to represent the response profiles(s) arising out of each respondent-cum-supervisor combination as per the planned ‘design lay out’. Two situations (blinded and unblended submission of responses) are differentiated and corresponding data analysis techniques are discussed. Variance components are assumed to be known in the study.

3. Estimation of Nonsampling Variance Components Under the Linear Model Approach (2009). Statistics in Transition, New Series, Vol. 10, No. 2, pp 193 – 222.

Summary: The importance of nosampling or measurement errors has long been recognized. [for numerous references see e.g., the comprehensive papers by Mahalanobis (1946), Hansen et.al. (1961), Bailar and Dalenius (1970), Dalenius (1974)]. Attempts have been made for estimating components due to nonsampling errors. The work in this area starts developing surveys, specifically designed to incorporate features which can facilitate the estimation of non sampling components such as reinterviews and/or interpenetrating samples. However most of the survey designs so far developed, though few, are very complex in nature [Fellegi (1964, 1974), Biemen et al. (1985), Folsom(1980), Nelson(1974)]. Here, a very simple survey design as well as a simple estimation procedure have been developed for the purpose of estimating simple as well as correlated response variances, namely interviewer variance and superviser variance.

4. A note on optimum inclusion probabilities in WOR-sampling scheme based on super population model (2009). (with T.P Tripathi), Pakistan Journal of Statistics, Vol XX(X), pp 221 – 226.

Summary: This paper deals with the problem of obtaining a set of optimum inclusion probabilities{πi;i = 1,2,…….N},optimum in the sense of having smallest average(δ-model based) mean square(design based) of the Horvitz-Thompson estimator for the population total Y. Since π ,s are dependent on model

16

parameters,a near optimum solution based on estimates of model parameters have been proposed.

5. Some improved variance estimators from a bivariate non-normal population (2009). (with T.P Tripathi). Pakistan Journal of Statistics, Vol XX(X) pp 165 - 194.

Summary:Given paired observations {(xi ,yi); i =1,2,…,n} on two variables X and Y on a random sample s from some bivariate non normal population.This paper considers an improvement of the customary estimator of population variance.A mixture (i.e.,a weighted combination) of the customary estimator and a suitably chosen statistic t is proposed. It has been shown that under certain conditions ,the improvement has been shown to be significant.

6. Existence of the BLUE for finite Population Mean under multiple imputation (2008): Statistics in Transition – new series, Journal of Polish Statistical Association, Vol 9, pp 233 – 258

Summary: Missing values not only mean less efficient estimates because of reduction in the sample size ,but also mean that the standard complete data methods can not be immediately used to analyse the data. Imputation, single or multiple, is a compensatory method for handling non responses and takes care of the fact that once the values have been filled in, standard complete data methods of analysis can be used. Here in this paper ,using multiple imputation technique, an estimator for the finite population mean in the presence of unit non response has been proposed and the estimator so proposed has been found to be theBLUE.

A very general non-linear cost model has been developed and discussed in the presence of nonresponse and an optimal solution of sample size for a given number of imputations or of number of imputations for a given sample size has been determined..

7. Stochastic Modeling of Buying Behaviour of Indian customers (2008), Calcutta Statistical Association Bulletin, Vol 60, pp 111 – 112.

Summary:A collaborative research on a problem of common interest and of immediate concern which Hindustan Lever Limited (HLL)-a multinational company was facing or likely to face in the near future was undertaken at the Indian Statistical Institute, Kolkata. The problem was to explain the purchase behaviour of frequently bought branded consumer products using stochastic models .For this ,the panel data after being coded to ensure anonymities were supplied to the ISI and on the basis of available data, modeling of the buying behaviour was made. To begin with, some descriptive measures were calculated to understand the data and finally, Dirichlet multinomial model was used for explaining the buying behaviour of the customer in the soecific segment with respect to the specific group of commodities.Because of not-so-wide coverage of the data and not-well-validated assumptions on the underlying distributions,the result failed to reveal much of the consumer behaviour pattern.

8. Some Shrinkage-Type Estimators for Variances of Univariate Populations (2007): JISPS (Journal of Indian Society of Probability and Statistics) in 2007 issue.

17

Summary:Here in this paper,an attempt has been made to improve upon the usual estimator for the variance σ2 by bringing in the role of some other statistics in addition to the sampling variance.The paper also discusses Bayes estimators for the parameter(s) of one parameter and multiparameter exponential families in general and in particular,,Bayes estimators for normal population. It also considers Bayes estimators for the parameters of Gamma population.

9. Unbiased Estimation of a Finite population: Case of Multiple Indirect Identifiability of Population Units (with B. K. Sinha and S. Sengupta) (2006): JSTA (Journal of Statistical Theory and application), Volume 5, 1, pp. 81-90.

Summary: Considered is the problem of estimation of population size and population mean in situations wherein the survey statisticians do not have direct access to the ultimate population units. Instead, there are available intermediate reference units and a network connection through which the ultimate units are linked. We develop a general procedure to tackle this problem and illustrate it via an example.

10. Information Support for Planning and Rural Development (2004): IX, Journal of Economics, Volume IX, Vidyasagar University.

Summary : This paper discusses the need of information net work - both electronic and others, for planning of rural development. It also discusses how information should flow from top to the bottom-the ultimate users especially to the farmers in rural areas.

11. Statistical Enquiry in identifying some specific castes as OBC’S in West Bengal, India, (2000): Bharatiya Samajik Chintan, XXII, -21.

Summary: At the instance of West Bengal Backward Class Commission, a study was undertaken at the Indian Statistical Institute to address the issues of social, educational and economic levels of some selected communities of West Bengal. .Based on primary data, an attempt was made to make an objective assessment of the Socio-Economic Profile of the five communities .Several indicators were used in this study to identify the relative social position of the individual communities .The present paper is based on some of the quantitative findings from the report.

12. Evolution of Statistics in India (with J.K. Ghosh et. al.) (1997): International Statistical Review, Volume 67, 1, pp. 13-34.

Summary: This is a brief history of the evolution of official and academic statistics in India which focuses mainly on the period 1930 to 1960 but traces its origins in antiquity and recent history. We also comment on how statistics has continued to evolve since the 1960s.This is a history of both institutions and people,who built and shaped them, and of ideas.

13. An Econometric Model of Exploration and Exploitation of Hydrocarbon (with M. Pal) (1997): Journal of Quantitative Economics, Volume 13, No. 2, pp. 29-44.

Summary: An econometric model of exploration and exploitation of hydrocarbon for estimation of discovery and production costs has been presented in this paper.

18

This is one of the two approaches developed for the purpose and is based on certain econometric relationships estimated on the basis of time series data on cumulated values of relevant values like reserves, costs of exploration and development drilling etc.

14. Stochastic Modeling and Forecasting of Discovery, Reserve and Production of hydrocarbon with an Application (with J.K. Ghosh et. al.) (1997): Sankhyā, Series B, Volume 59, Part 3, pp. 288-312.

Summary: The primary objective of this paper is to present some results in connection with forecasting of discovery and production of hydrocarbon in a partially explored basin. The focus in this article will be on stochastic modeling of the purpose of discovery, based on the idea of subjective probability or superpopulation. The results of a particular basin in India have been presented here.

15. On Minimax Allocation of Stratified Random Sampling when only the order of stratum variances is known (with M.Pal) (1994): Statistics and Decisions, Vol. 12, pp. 195-201.

Summary: This paper proposes the minimax criteria for obtaining the sample sizes to different strata when only the ranks of the stratum variances, apart from the stratum sizes, are known and obtains a very simple and elegant solution to this problem

16. Estimating unknown dimensions of a binary matrix with applications to estimation of the size of a mobile population (M.Pal and B.K. Sinha) (1992): Statistics and Probability Edited by S.K. Basu and B.K. Sinha.

Summary: For two fixed and positive integers N and T, Let ( ))( ija∆ be a matrix of order TN × with elements saij assuming values 0 and/or 1 with the restriction that for every i.

NiaM ij

T

ji ≤≤≥= ∑

=

1,11

However,

TjaN ij

T

ii ≤≤≥= ∑

=

1,01

With

∑∑∑∑ ≥== NaNM ijji

We considered a situation where T is known in advance but N is unknown. The problem we address is that of estimation of N unbiasedly. We propose to develop a reasonable theory for this problem after formulating the same in the frame work of Finite Population Inference.

19

17. Some results on 1T class of linear estimators (1998): Jour. Indi. Society of Agri. Statistics, Vol. LX, No. 1, pp. 1-8.

Summary: 1T class of linear estimators is examined to obtain a biased sub class of estimators better than the sample mean y .

18. Estimation of Lorenz Ratio from a finite population (with M.Pal) (1988): Anvesak, Vol. 18, Nos. 1 -2, June – December, pp. 29 – 62.

Summary: To provide an unbiased estimator for Lorenz Ratio without assuming any distribution of the population, a sample-theoretic approach has been taken to estimate Lorenz Ration from a finite population under a general sampling design. It has also been observed that the usual estimators for LR are biased under simple random sampling.

Also an effort has been made to provide an estimate of Lorenz Ration in case of rank of thi observation is known. Uusing this knowledge of rank, two alternative estimators have been proposed under PPSWR – scheme.

19. On some estimates of poverty measures (with M.Pal) (1988): Calcutta Statistical

Association Bulletin, Vol. 37, March and June, Nos. 145 – 146, pp. 81 90. Summary: There are now a number of poverty measures available in the

literature. Some of the measures are alternative to each other and some claimed to be superior in some sense to many others.

When significant work has been done in developing the alternative measures, not much attention has been paid to the problem of estimation of these indices. Estimation does not pose very serious problems in the large sample, but when are deals with a small sample, which may typically be the case in reality, situations become quite different. In fact usual estimators become biased for some of the indices. In this paper, alternative estimators for these cases have been proposed. Other properties of the estimators and some other relevant issues have also been examined.

20. A Unified Approach to estimation of Lorenz Ratio from a finite population (with M.Pal) (1988): Sankhya, Series B, Vol. 50, part 2, pp. 215-223.

Summary: Most of the results relating to estimation of Lorenz Ration (LR) are based on the assumption of some distribution of the population. Not much attention has been paid in the literature to provide a design based unbiased estimate of LR. To the best of Author’s knowledge, Taguchi’s (1978) is the only one which however can hardly be used in practice. In case Y , the population mean is known, an effort, in this paper, has been made to provide some unbiased estimator of LR together with estimates for the variance of estimators. The relevant expression have been found under the general sampling design and then, in particular, the cases of SRSWR and SRSWOR have been discussed.

20

21. Use of Prior information on some parameters in estimating population mean (with T.P. Tripati, S.D. Sharma) (1983): Sankhya, Series A, Vol. 45, part 3, pp. 372-376.

Summary: We consider the problem of estimating population mean Y of a character y, using information on some other parameters of y . A class of estimators which are linear functions of y and a suitably chosen statistic t is presented. General properties of this class are studied and the optimum weights and the resulting optimum mean square error is found. A general technique of generating estimators better than sample mean y and Searl’s estimator (1964) is given and a number of such biased estimators are identified for some choice of t, under very moderate conditions depending on the prior knowledge of the quantities which are smaller or greater than the actual values of some population parameters.

22. A Note on the Estimation of Mean Square Error(1982): The Aligarh Journal of Statistics, Vol. 2, pp. 38 -40.

Summary: In an earlier paper, Maiti and Tripathi (1981) obtained a biased subclass in 2T -class of linear estimators, where the well known Horvitz-Thompson estimator fails to be better, better in the sense of having smaller mean square error than one belonging to the biased subclass, of course, under some moderate conditions. In this paper, we present the conditions under which the estimates of mean square error will be non-negative.

23. Some 2T - class of estimators better than H-T estimator (1981): The Aligarh Journal of Statistics, Vol. 1, No. 1, pp. 52-58.

Summary: The 2T class of estimators for the population total of a character y, in case of general sampling design is revisited and a subclass of biased estimators from 2T better than H-T estimator THY −

ˆ is identified. It is found that in case of a class of sampling designs, we may always generate estimators better than THY −

ˆ .

We also study another biased subclass of estimators iisi

pyT /*2 ∑

= λ where

Nixxp i

N

iii ,,2,1/

1== ∑

=

, x being an auxiliary character and λ is a suitably

chosen constant. Some members from *2T are shown to be better than THY −

ˆ under a super population model.

15.2.Abstracts of some of Research Papers Appeared in the Proceedings of the Conferences: 1. A Generalised Horvitz-Thompson Estimator for Population Total with an Application

in estimating number of species in the Region of Silent Valley in India (2015):

21

National Conference on Recent Advances in Statistical and Mathematical Sciences and their Applications (RASMSA 2014), Kumayun University, India, October 2014. Abstract: considered is the problem of estimation of a population parameter, say total, where the survey statisticians do not have direct access to the ultimate population units, as may be the case of a mobile population. On the other hand, there may be some intermediate reference units and a well defined unique net work connection through which the ultimate units may be reached. Making use of such intermediate reference units, a Generalised Horvitz-Thompson estimator has been defined and deployed in estimating number of species of a real life population.

2. Estimation of Measurement Variance in the Context of Environment Statistics (2014) 2nd ISM International Statistical Conference 2014 with applications in Science and Technology (ISM-II 2014), Pahang, Malaysia, August 12 – 14, 2014. Abstract: Timely, reliable and comparable data are needed for any statistical analysis. The field of environmental statistics has no single overreaching internationally agreed classification for statistical purpose. Lack of proper and uniform definitions, unambiguous classifications pose serious problems to procure qualitative data causing measurement errors – errors in coverage as well as in technique. For example, for obtaining statistics related to an area under a crop, perimeter bias causes not to have accurate information. We consider the problem of estimating measurement variances, so that some measures may be adopted during the field work to improve upon the quality of data on environmental goods and services and on their valuation in economic terms, which is necessary for policy formulation to achieve overall sustainable development of the society. The measurement technique considered here is that of employing personal interviewers and sampling design adopted is that of a two-stage sampling.

3. Estimation of Variance Horvitz-Thompson estimator in the presence of measurement error: The linear model approach. 1st International Conference on Information Management and Statistics (ICIOMS) during September 1 to September 3, 2013, Kuala Lumpur, Malaysia. Abstract: The importance of identifying and finding ways and means to control the non-sampling errors has long been recognized [Mahalanobis (1946); Hansen etal (1951), (1961), Bailer and Dalenius (1970), Dalenious (1974)]. But not much work compared to controlling sampling errors has progressed so far. Attempts have been made here for estimating the two components of the variance, namely sampling variance and measurement variance under the linear model approach. Primary need to estimate measurement variance is to obtain repeat measurements and to follow the principle of randomization i.e., an appropriate survey design needs to be developed. In this paper, a survey design based on Symmetric Balanced Incomplete Block Design (SBIBD) has been developed and deployed in estimating non-sampling variance component of the Horritz-Thompson H-T estimator. The sample design adopted is that of cluster sampling and the estimator used is the H-T estimator.

22

4. Interactive linear model in survey sampling: Appeared in the conference series as well as in conference volume of 7th Annual Conference on Statistics during 17 – 20 June, 2013, 2013, Athens, Greece. Abstract: Considered is a linear interactive model in the context of survey sampling. The situation arises when investigator and/or supervision interactions are contemplated in the responses. Blinded situations have been discussed here in.

5. Some Aspects of Decentralised Planning International Conference on Recent Advances in Mathematical Statistics and its Applications in Applied Sciences in Collaboration with Indian Statistical Institute, Kolkata; Department of Statistics, Gauhati University DST FIST and UGC(SAP) DRS-I, Department, Gauhati, Assam, India, December 31 2012 – January 1-2, 2013.

Abstract: A seminal talk on the theme of the decentralized planning was delivered. It discussed the decentralized administration, additional responsibility assigned by 73rd amendment of the constitution, India for rural administration and need for strengthening local Government, which inturn needs the Development of Statistical Information System (SIS). The SIS discussed has specified such major components as (i)items of information needed for local level planning, (ii) designing input output formats amenable to computerization, (iii) method of data collection, collation, and (iv) software developed for creating RDBM (Relational Data Based Management) and for processing the data.

It also discussed how a consensus development plan for one Gram Panchayat could be developed on the basis of information created by the SIS.

6. A tale of two Externally Funded Project Reports: The MP Experience, Pulakesh Maiti, J.K. Ghosh and B.K. Sinha. Presented at the conference of Central and State Statistical Organsation (COCSSO), April 10-11, 2007, New Delhi – 110012. A rudimentary Statistical System began in Indian during the Hindi-Budhist period. It evolved into a more mature system when the Moghuls ruled India. A rapid growth took place during the British Period. Further growth and modernization with focus on the country’s socio-economic programmes occurred after India became independent in 1947. Currently, the system has some problems, but also it has, on its record, many achievements and much promise. An appraisal of the present Indian Statistical System suggests some improvements needed in the statistical framework both at the state and at the centre. Report on National Statistics Commission (August, 2001) has emphasized, among other issues, on the adequacy, reliability and timeliness of the information generated by the system. The State Statistical Bureaus (SSD/DES/BAES) have been required to be strengthened with respect to the above goals.

At the same time, the 73rd and 74th amendments of the constitution have squarely laid the responsibility of local planning to local bodies. Local level planning will need local level statistics with respect to decentralized planning. The Commission has also pointed out that I the future, the new self-managed local bodies in cities and rural areas created by virtue of the above two amendments, will be in a position to supply lots of data.

Evaluation of the performances of two externally funded projects, viz., one ISI-PWI Project (1999) and the other JP Associates (P)-Project (July, 2006) have been made, keeping in mind the present need from the existing systems. It may be observed that the project undertaken by JP-Associates has mainly focused on identifying specific requirements for strengthening

23

SSB;s and, hence, no attempts have been made on identifying items of information needed for local planning. Although, the primary concern of the ISI-PWI project was to build up the Statistical Information System (SIS) for local level planning, with respect to decentralized planning, it has also investigated in to the strength and weakness of the Directorate of Economics and Statistics (DES) of the then undivided M.P.

In effect, the work involved in the ISI Project is two- fold: Development of SIS for Strengthening Local Government in M.P. as well as strengthening the DES, though in the later domain, it is not as detailed as in the report of JP Associates. The presentwrite-up has made a comparative study of the contributions from the two projects with respect to both strengthening SSB’s and Local Government in undivided M.P. . It has also provided an out line of the volume of work, cost/time frame work for generating data base for basic statistics for local development.

7. Estimation of a Finite Population Total: Interactive linear models in Survey

Sampling, Bikas K. Sinha , (with Pulakesh Maiti); An invited talk: International Conference on Statistical Data Mining for Bio-informatics, Health Agricultural and Environmental: 21 -24 December. Proceedings, Department of Statistics, University of Rajshasi, Bangladesh.

Abstract: Considered is the set up of simple i.e., direct response on a quantitative response variable Y in the context of a finite lablelled population of size N. In actual surveys, it so happens that we need investigators and often supervisors as well. We depict a situation where in there are possibilities of investigator effect and/or supervisor effect on the response profile finally received by the data collecting agency. Of course, there effect may be assumed to be random, having mean zero, non-interactive within and between the two sets of people. The problem is to unbiasedly estimate the finite population total of the response Y by incorporating a fixed size (n) sampling design and by administering the sampling design in a situation where the above two types of random effects are likely to be present.

8. Information created and developed for Decentralised planning to two different blocks

of the district of Howrah, West Bengal (2012), Pulakesh Maiti, Proceedings of International Workshop on large scale National Surveys, pp. 20-28, Department of Statistics, University of Rajshahi, Bangladesh, ISBN 987-984-33-7470-7. Abstract: To help one Gram Panchayat (GP) prepare a data based local level plan on the basis of the principle of Decentralisation, discussions have been made on how the information collected on items as listed in 11th and 12th schedules of 73rd and 74th Amendment of the constitution were of great help in identifying most needy areas with respect to (i) infrastructure facilities, (ii) health, (iii) education, (iv) drinking water facilities, (v) rural electrification, (vi) other living conditions, namely daily marketing facility, availability of medical shops etc., which would require immediate attention for development in the areas. As an illustrative example, two blocks of the district of Howrah, West Bengal, India have been considered – one for from the city and the other being adjacent to the city.

24

9. The National Sample Survey Organization (NSSO) and the quality of its data – A Brief Review (with J. K. Ghosh) (2004). Presented as an invited paper at Joint IMS-SRMS meet on current trends in survey sampling and official statistics, Calcutta, India, January 2-3, 2004. Appeared on the web as University of Maryland/IMS/Papers Abstract: After a description of the history and salient features of the surveys undertaken by the NSSO, we focus on quality of its data in recent times. We summarize some common criticisms as well as a couple of analytical studies in the literature. We also present our tentative view that except Annual Survey of Industries (ASI), both thin and large rounds provide reliable data.

10. The Indian Statistical Systems at Cross Roads – An Appraisal of Past, Present and

Future (with J. K. Ghosh) (2004): An invited paper at Joint IMS-SRMS meet on current trends in survey sampling and official statistics, Calcutta, India, January 2-3, 2004.Appeared on the web as University of Maryland/IMS/Papers. Abstract: A rudimentary statistical system began in India during the Hindu-Buddhist period. It evolved into a more mature system when the Moghuls ruled India. A rapid growth took place during the British period. Further growth and modernization with focus on the countries socio-economic progress occurred after India became independent in 1947. Currently it has many problems but also many achievements and much promises. We traced the history of the system and end with an appraisal of its present status and future prospects and much promises. We traced the history of the system and end with an appraisal of its present status and future prospects.

11. The 275th Indian Association of Special Library and Information Centre (IASLIC)

study circle lecture “Information System on Decentralized Planning for Rural Development”, Presented as an invited lecture on June 17, 2004.Proceedings of the Conference of the Golden Jubilee Celebration ,September 2004-August,2005. Abstract: Planning for development involves four different types of activities, formulation, implementation, monitoring during implementation and evaluation on completion. To carry out each of these activities, relevant, reliable and timely information is needed at every stage. The above presentation makes an effort to determine the items of information needed for strengthening rural decentralization

12. Information Support for Planning and Rural Development :Key-Note Address Proceedings of the Conference on “National Seminar on Information Support for Rural Development” organized by Dept. of Library and Information Science, Vidyasagar University: Nov. 18-19, 2003: Abstract: Information have been collected and used in the Indian subcontinent from antiquity, but major changes in collection and use took place during the British period (1757-1947) in Indian history. New imperial needs dictated some of the changes, but much of it took place indirectly as a result of western education and a spirit of scientific curiosity and experimentation. Interest in rapid social, economic and technological development changed the face of information need of the country and

25

added a new dimension to information system after India’s independence in 1947, especially in planning rural development.

15.3. Papers under Communication:

1. Revised version of the Indian Official Statistical System Revisited by P. Maiti., T.J. Rao and J.K. Ghosh: Submitted to Sankhya Series B. Abstract: We first discuss some emerging problems and then focus on what seems to be most important as well as puzzling among these problems, namely the divergence between survey based estimate and estimate based on national accounts. A couple of simple implementable strategies have been proposed to capture sparsely scatteredness of the informal sector, which is suspected to be the most important single factor causing the aforementioned divergence.

2. Mean Square Error Decomposition Model under Two-Stage Sampling: Estimation of Population mean by Pulakesh Maiti and Mahendra S. Shiton. Submitted Statistics in Transition, Poland. Abstract: The problem is to estimate unbiasedly the finite population total/mean of a response variable Y by incorporating a two-stage sampling design and by administering the the sampling design in a situation where the intervention due to the investigators are likely to be present. An estimator for the population mean has been defined taking care of investigator’s effect. General expressions for measurement variance as well as sampling variance have been provided considering the mean square decomposition model.

3. Estimation Measurement Variance Under Two-Stage Sampling by Pulakesh Mait, Communicated to Statistics in transition, New Series, Poland. Abstract: Considered is the mean square error decomposition model for estimation of population mean and the problem of estimation of measurement variance under two-stage sampling. The measurement technique considered here is that of employing personal interviewers for data collection. The sample mean has been defined taking care of interviewers effects and the measurement variance associated with the estimator has been derived with an effort to estimating the measurement variance.

16. Book Chapters:

1. Interactive Liner Models in Survey Sampling (2014): Essays on Mathematics

and Statistics, Edited by Codrufa Stoica, Athens Institute for Education and Research (ATINER), pp. 141 – 145.

26

2. Indian Statistical Institute – Numbers and Beyond (1931-1947) (with J. K. Ghosh and Anil Bera) (2007): Appeared in Project of History of Indian Science, Philosophy and Culture (PHISPC) Volume 1, “Science and Modern India: An Institutional History C ‘1784-1947’ edited by Professor Uma Dasgupta.

3. Development of Statistical Information System (SIS) for Decentralized Planning

(2004), pp. 198-248: Participatory Decentralized Planning in India: Issues of Finance and Statistics Information,Edited by Professor Sau , FIRMA KLM Publication, Calcutta.

17. List of Projects:

1. A Collaborative Research Project (with Rajshahi University, Bangladesh)

entitled Study of Quality Education in Some Districts of Bangladesh, (2012). 2. The externally funded project entitled Diagnostic Study of the Registered

Closed Units (2012-13), Indian Statistical Institute, Calcutta. Nature of Participation: Principal Investigator

3. The externally funded Project on Socio-Economic Conditions of five minority communities in West Bengal (2008 -10 ).Indian Statistical Institute, Calcutta. Nature of Participation: Principal Investigator.

4. The externally funded Project on Development of Data Base for Decentralized Planning in Howrah District (2006 - 2007). Nature of Participation: Member Statistician

5. ISI-HLL Collaborative Project on Business Research (October 1999-August 2000 : Indian Statistical Institute, Calcutta, India. Nature of Participation: Co-Principal Investigator.

6. ISI-PWI Project on Strengthening Local Government in Madhya Pradesh, India (May 1998- April 1999). Indian Statistical Institute, Calcutta and P.W.I., Calcutta. Nature of Participation: Co-Principal Investigator.

7. Mid-Term Review of IPP-VIII in Calcutta Metropolitan Area (1997-1998): Indian Statistical Institute, Calcutta, India. Nature of Participation: Member Statistician.

8. The Socio-economic-demographic and cultural pattern of the female labour force participation of the North West and The Cape (June 1995-December 1996): CENSTAT, HSRC, South Africa. Nature of Participation: Principal Investigator.

9. Survey of family and community life in the selected communities of The Cape Peninsula of The Republic of South Africa (January 1995-December 1996): CENSTAT, HRSC, South Africa.

Nature of Participation: Principal Investigator.

10. Community attitudes and preferences pertaining to cemetery and cremation related issues in the East Rand in Republic of South Africa (May 1995-December 1996): CENSTAT, HSRC, South Africa.

27

Nature of Participation: Principal Investigator.

11. Identification of Other Backward Classes (OBC) in West Bengal (1994): Indian Statistical Institute, Kolkata.

Nature of Participation: Member Coordinator

12. Evaluation of Total Literacy Campaign (TLC) in West Bengal (1990 - 91): Indian Statistical Institute, Kolkata.

Nature of Participation: Member Statistician.

13. ISI-ONGC Collaborative Research Project (1985 – 1990): Indian Statistical Institute, Kolkata.

Nature of Participation: Member Statistician.

14. On Domestic Tourists survey in Orrisa (1988 – 89): Indian Statistical Institute, Kolkata. Nature of Participation:Member Coordinator

15. On Medical Facilities available in Calcutta Metropolitan Development Area (CMDA) (1977 – 78): Indian Statistical Institute, Kolkata.

Nature of Participation: Project Assistant.

16. Calcutta Urban Poverty survey (1976): Indian Statistical Institute, Kolkata. Nature of Participation: Project Assistant

17. Cost Benefit Analysis of Rural Electrification (1976 - 77): Indian Statistical Institute, Kolkata.

Nature of Participation: Project Assistant

18. On the study of Job Oriented Education based on Regional Demand survey (1975 - 76): Department of Mathematics, Indian Institute of Technology, Kharagpur, West Bengal, India.

Nature of Participation: Research Scholar. 18. Project Reports (1975-76 - ):

1. The externally funded project “Diagnostic Study of Registered closed units: (2012-2013). The Indian Statistical Institute. Nature of Participation: Principal Investigator Summary: The study was to identify and classify registered closed units by such reasons of closures as (i) financial, (ii) administrative and (iii) marketing. The Indian Statistical Institute under the leadership of Dr. Pulakesh Maiti have devised the project and used the expertise in analyzing the data including (i) designing and testing of the questionnaire; (ii) training programme for the investigators; (iii) supervision of the field work; (iv) data entry and validation; (v) development of the tabulation programme, and the appropriate estimators;

28

(vi) preparation of the Report.

The findings have helped the Government specify the different causes of the closures of the units with their identification the government has also identified amounts of lands under closed units, which could be utilized for other Micro and Small Scale Enterprises who were eager to make the industrial venture.

2. A Collaborate Research Project ‘Study of Quality Education in some Districts of Bangladesh (2012--) Nature of Participation: A member Statistician Summary: The study has such specific objectives as to (i) observe infrastructure and environment of the schools; \ (ii) examine some essential aspects of quality education; (iii) identify the success and short comings and examine all the relevant issues

and constraints in the attainment of quality education.

3. The externally funded Project on Socio-Economic Conditions of five minority communities in West Bengal (2008 - 2010).Indian Statistical Institute, Calcutta.

Nature of Participation: Principal Investigator. Own Activity:

The sampling design developed was that of a three stage stratified sampling one with ultimate sampling units being the households. The excersises on arriving at the above sampling design and of determination of toal sample size and their allocation into different hierarchical units were carried out by myself. Representation of sampling units from rural as well as urban areas of the district has been made possible through the above design. The findings of the study helped the Government make assessment of the socio-economic condition of the minority communities and take appropriate measures from the view of development policies.

4. The externally funded Project on Development of Data Base for

Decentralized Planning in Howrah District (2006 - 2007). Nature of Participation: Member Statistician Summary: The terms of reference of the project was as follows: (a) Designing of formats for data collection and collation;

29

The variables should be identified and assimilated into formats keeping in view that each of them should have properties like (i)easy to understand, (ii)concise, (iii) amenable to computerization;

(b) Imparting training to the investigators; (c) Conducting the sample checks of the data.

Two days of classroom training program on 28-02-07 and on 01-03-07 were preceded by five days of pilot study during the field work of March 2007. An intensive scrutiny program was developed and the data were critically examined. Information thus created for decentralized planning helped the District Planning Committee of the District Howrah for local level planning. In fact a data based local level plan was developed for one Gram Panchayat Amardaha of the district. The data base created under this project helped the district planners identified the vulnerable areas with respect to (i) informative facilities on health, education, (ii) drinking water facilities, (iii) rural electrification and other living conditions namely daily marketing facility and availability of medical shops etc.

Own Activity: I was responsible for designing the schedule in English, which got translated later on in Bengali by some other team members in order to facilitate data collection job in the villages of the distict. Trainings were imparted to the investigators by myself and a pilot study was undertaken under my supervision to finalise the format of data collection and to estimate the manpower needed. With the help of the scrutiny schedule developed by myself, data were critically examined through an intensive scrutiny program. I did all sorts of lianson activities—academic as well as administrative, with the district officials as well as with the elected functionaries of the Zilla Parishad. A number of meetings were held in the offices of the District Magistrate and of Sabhadipati to finalise the data base. Information this created for decentralized planning helped the District Planning Committee for smooth running of government at the local level. Data base created helped the policy makers identify the areas which would need immediate attention for planning.

5. Indian Statistical Institute–Hindustan Lever Limited (ISI–HLL) Collaborative Project on Business Research (October 1999 – August 2000): Nature of Participation: Co-Principal Investigator. Summary: There are many models of buying behaviours available in the existing literature on Marketing Science. Among these, the Ehrenberg Bayesian model seems to have given good results. One area of interest which was important to the HLL was to this Ehrenberg model, if this could be applied to consumer panel data in Indian Context. For this, we needed to study the validity of the Ehrenberg

30

model in Indian cross sectional data. The appropriateness of the distributional assumptions of the model namely, (i) assumption of the negative binomial distribution on the number of purchases

and (ii) assumption of the Beta priors on pj , the choice probability of the jth brand

in a product group were examined for a given product. The study helped the company understand the extent of demand of the goods; it also helped the company understand the buying behavior of the customers and the market share of the different commodities. This also helped the policy makers arrive at the appropriate decisions.

Some Technical novelties of the work were as follows: (i) The data used for the purpose of analysis consisted of, among others, the

information on household identification number, brand code, month number and quantity of the product purchased. The number of purchases was the basic input of the model. Thus for the model verification exercise, the data on the quantity (in grammes) purchased of each brand by the households had to be converted into the number of units. This requires a knowledge about the standard size of each brand. In the absence of such information precisely and also for the sake of rendering flexibility to our analysis, we used, alternative sizes, viz, 125 grammes, 250 grammes, 500 grammes and 1000 grammes with proper rounding off to obtain the number of units. By this procedure, we generated four sets of data on number of purchases of each brand made by each household in each of the months. The data on number of purchases based on 1000 grammes was in agreement with the experiences as realised by the practitioners of the HLL.

(ii) “Empirical Bayesed” as well as “Hierarchical Bayesed” estimates for the model parameters were obtained;

(iii) Data failing to support the assumption of negative binomial distribution on the number of purchases, insisted us to make some further studies, for each of income group separately, as if, the data resulted from a mixed distribution with some mass p at n = 0 (n being the number of purchases) and a binomial of (3, 1-p);

(iv) It was interesting to observe that there had been a good fit on the distribution of the number of brands. The distribution appeared to follow a truncated geometric distribution (truncated at Zero);

(v) Assumption of Beta prior on pj, the choice probability for the jth brand also appeared to be not tenable in the Indian Data Context;

(vi) The reasons behind the model for not being fitted to the Indian data were tried to be found out. One of the reasons was that unlike the European family, the number of decisions makers in an extended family in a country like India may be many. The model was revised by introducing into it a new stochastic variable namely, the number of decision makers at the household level to examine if the revised model was supported by the Indian data.

31

(vii) The households were classified according to family size and income class and the buying behaviour across such classes were examined to detect any patterns of similarity.

Nature of my work: The above exercises were carried out by myself along with my other co-team

members especially with Professor Manoranjan Pal. In addition to the work made and indicated in the earlier section, as the co-

principal investigator, I had to liason with Dr. Siddhartha Roy, from HLL. This involved a lot of administrative work. In particular, maintaining project account was also one of the major responsibilities entrusted to me.

6. Indian Statistical Institute – Price Water House India (ISI - PWI) Project on

Strengthening local Government in Madhyay Pradesh, India (May 1998-April 1999):

Nature of Participation: Co-principal investigator

Summary: Planning for development involves four different types of activities, formulation, implementation, monitoring during implementation and evaluation on completion. To carry out each of these activities, relevant, reliable and timely information is needed at every stage. The 73rd and 74th Amendments of the constitution made by the Government of India have squarely laid the responsibility of local planning on local bodies. It stands to reason that information needed for such planning should also be collected and managed by local bodies and hence the Government of India had requested Asian Development Bank (ADB) for technical assistance to strengthen local Government of Madhya Pradesh as part of the reform agenda of the state. At the request of Price Water House India, who was one of the prime contractor to the Bank for this purpose, an inter-firm agreement between PWI and ISI was made to develop the Statistical Information System (SIS) as a part of the project assignment. The SIS was envisaged to be a statistical database for rational decision making. It was expected to address the information needed for planning at Panchayat, Janpad, district and higher levels. This was developed in accordance with the 73rd /74th amendment of the constitution of India, 1992. The statistical information system which comprises of such major components as (i) Computer hardware forming he continues of statistical information; (ii) Computer software to process the information; (iii) Statistical data, the actual context of the system; helped to make resource mapping of a given area, which in turn, helped develop a development plan with respect to the principle of decentralized planning.

32

Highlights of some of the Technical Features: (i) Extensive surveys of the areas of activities as listed in the 11th schedule of 73rd

and in the 12th schedule of 74th amendment and of the lists of items prepared by the expert committee on small area statistics and also of the report by Hanumanth Rao Committee were made.

(ii) A survey on identification of the availability of required information with an analysis of the existing data gap was made.

(iii)Twelve (12) rural and thirty one (31) urban schedules with indications of respective sources of data serving as SIS input manual were developed.

(iv) The appropriate method of data collection was suggested. (v) Designing the output format for (a) general information on variables

considered for decentralised planning, (b) quarry based information, (c) report based information and for (d) summarised information, was made.

(vi) The following two research papers have been prepared out of the above project. (a) Development of Statistical Information System (SIS) for decentralised

planning by P. Maiti. (b) Information Support for planning a development – with reference to

Rural Development by P. Maiti.

Own Activity: The work involved in developing the SIS consisted of (a) identification of required data items, (b) designing of formats of data collation, collection and compilation, (c) specification of output formats, amenable to computerized data base and (e) organization of workshops in collaboration with the state government to finalize the methodology of data collection and data formats. I was stationed at Bhopal, Madhya Pradesh and executed the above work during 1998-99. The SIS developed for decentralized planning has the following major components: (i) Computer hardware forming the container of Statistical Information; (ii) Computer software to process the information; (iii) Statistical data, the actual content of the system;

In addition to the work as mentioned in the “Highlights of Some of the Technical Features”, the following works need special attention. Estimation of Work Load:

The volume of information for each village was calculated and appeared to be approximately 8000-10000 alphanumeric characters or bytes and for each urban local body (ULB) about 10000-15000 bytes. The total volume of information for the sate as a whole was about 1000 million alphanumeric characters or one Gigabytes (GB). The total number of documents was 72000 for the rural sector and only about 400 for the urban sector.

33

It was estimated that to collect the information level and prepare a document for a village, 3 mandays would be needed and for each document in the urban sector 6 mandays would be needed. To key in the data on to floppies, the total input volume was thus about 1 GB. At the rate of 5 KB per hour, a total of about 100 man-years of data entry effort was required. For the work, I had to stay at Bhopal most of the time during the project period.

Strengthening Local Government in M.P., PWI & the Award : The project was a part of the Technical Assistance provided to the Government of Madhya Pradesh (MP) by the Asian Development Bank through Price Water House India. On behalf of the PWI, the Statistical Information System was developed by the Indian Statistical Institute, Kolkata through ISI – PWI collaborative research project. The training component of the project was assigned to some other agency.. Some of the video programmes produced from the project report by the PWI has bagged SONY/ICD – 1999 Award in Japan.

34

A Description of the work involved in development of SIS METHODOLOGY

Additional Responsibility Assigned by 73rd & 74th Amendment of municipality Act and Panchayati Raj Act

Assess data Requirements

Match with Existing Data (Raw & Published

Gap Analysis

Data Collection Scope of System Client Requirement- Approval Feasibility & Select from Existing Data Data Formatting for Computerisation Scope of Design Computerised Implementation Database Test & Deliver Client Approval Training

If any new data Take on data collected

Pilot Implementation at HO/Bhopal District Role on Implementation Plan Client Including Data Take on of Approval Remaining & New data

Project Completion Review

35

7. Mid term Review of Indian Population Project VIII in Calcutta Metropolitan area (CMA) (1997-1998):

Nature of Participation: Member Statistician. Summary: The objective of this project was to study on different facets of IPP-

VIII. One important component was to assess the contributions made by IPP-VIII Project in spreading family limitation practices and in improving maternal and child health care among the beneficiary households. Several satellite enquiries were also conducted to assess (i) the role, capability and motivation of IPP-VIII personnel, namely those of health officers, chairpersons, municipalities, (ii) the effectiveness of training programmes and IEC programme organised by Calcutta Metropolitan Development Authority (CMDA), and (iii) the qualitative and quantitative progress of civil construction works undertaken for the project IPP-VIII. Finally, evaluations on health facilities such as the physical ones with their mentainance and repair, on the availability of drugs and equipment and on the staffing pattern were made.

On the basis of the findings of the study on “assessment of IPPVIII project in spreading family limited practices and in impressing material and child health care”, government took the appropriate measures on extending health facilities such as physical ones with their maintenance and repair, on the availability of drugs and equipment and on the staffing pattern etc. The followings were some of the technical novelties. On estimates of some of the non-sampling errors: (i) An estimate of the over coverage error due to inclusion of non-population units because of imperfections in the frame was found to be around 15%; (ii) The rate of non-availability of the responding beneficiary households was found to be around 3.7%

On Sampling Design: In developing the sampling design, the following considerations were made.

(i) All the programmes in Health Administrative Units (HAU’s) did not start functioning at the same time. Some started before April ’95 and some, since April ’95. Therefore a stratified multistage design was adopted, stratification being made on the basis of the time of start of functioning of the different programmes. Simple random sampling was adopted at different hierarchical stages, namely at HAU’s, at subcentrers, and at blocks, except at the ultimate stage of selecting the beneficiary households, where linear systematic sampling was used.

Because of the variation in fertility among married couples with respect to their ages, before selecting the sample households, all the households in the selected sample blocks were further classified into four sub-strata on the basis of presence of a married couple and the age of the wife in the household. The required number of sample households from each sub-stratum was selected by systematic sampling. Thus a multistage stratified

36

sampling with further deep stratification was developed for selecting the ultimate respondents.

(ii) Since the project aimed to make an assessment of the impact of IPP-VIII programme, the sampling design should have been repetitive in nature at least over two periods to produce longitudinal data. But since this was not possible, the problem was taken care of by properly designing the questionnaire, where the respondent was asked during the course of data collection to produce information on present overall demographic characteristics, attitudes and practices of family planning along with those within five years from the date of the survey. Since information related to vital events like births and deaths within the family, not much recall lapse error was reported. The error due to recall lapse was within control.

Estimation Procedure: i: Subscript for a block, t: Subscript for a stratum, j: Subscript for a household, k: Subscript for a sub-centre, l: Subscript for a HAU, H(U): Number of HAUS in the u-th stratum (u= 1,2). h(u): Number of sample HAU’s selected from the uth stratum (u=1, 2).

)(ulS : Number of sub-centres in the lth HAU of the uth stratum (u= 1, 2). ( )uls : Number of sample sub-centres selected from lth HAU of the uth stratum.

)(ulkB : Number of blocks in the kth sub-centres of the lth HAU in the uth stratum.

)(ulkb : Number of sample blocks selected from the kth sub-centres of the lth HAU

in the uth stratum. T: Total number of sub-stratum ( T=4).

)(ulkitN : Total number of households in the tth substratum of the ith block selected

from the kth subcentre with lth HAU in the uth stratum. )(u

lkitn : Number of sample households selected from the tth substratum of the ith block selected from the kth subcentre with lth HAU in the uth stratum.

)(ulkitjY : Value of the study variable for the jth sample household in the tth

substratum of ith block selected from the kth subcentre of lth HAU. Yu: Total value for the uth stratum ( u = 1, 2) Unbiased estimates of Yu and Y were computed as ∑∑∑∑∑=

l k i t j

ulkitj

ulkitju YWY )()(ˆ

where, lkit

lkitu

lk

ulk

ul

ul

u

uu

lkitj nN

bB

sS

hHW ... )(

)(

)(

)(

)(

)()( = is the multiplier associated with the value

)(ulkitjY and )2()1(

ˆˆˆ YYY += .

37

Nature of my work: In addition to above work executed by myself in developing sampling design,

estimation procedure and some error analysis as outlined above, special mentions need to be made on my following work also.

(i) On selection procedure: Simple Random Sampling (SRS) was adopted at all the different hierarchical stages except at the ultimate stage,where linear systematic sampling was employed to select the ultimate stage units.

(ii) The stratified multistage design as described above was developed and deployed to collect data, and consequently, the multipliers were calculated.

(iii) Combined use of Probability and Non-Probability Samples: In addition to the probability sample of the beneficiary households which constituted the population under study, use of non-probability/purposive selection of chair persons of municipalities and knowledgeable persons were made for their assessment and perceptions of health functionaries who were entrusted with the job under IPP-VIII population project.

(iv) On Questionnaire Design: Mid term evaluation of the IPP-VIII project was meant to make an assessment of its impact on the beneficiary households. For this purpose, the sampling procedure should have been repetitive in nature to produce longitudinal data for comparison. But this was not possible. As an alternative, the problem was taken care of by properly designing the matching schedules in such a way that the respondents might provide on their present overall attitude on family planning practices along those within five years from the date of survey. Questionnaire design was also made to avoid possible “recall-lapse error” due to the time lag.

(v) Appropriate measures were taken on the choice of investigators to reduce the gender bias.

(vi) The following exercises were made on the “item non-response” for the items, namely “effectiveness of teaching”, “use of text books”, “use of teaching aids” and use of “Black Board”.

Table 7A: Percentage distribution of item response indicating “Effectiveness of Teaching” (IPP VII Project , 1997-98):

Items Response Non-response Yes No Sufficient time to take notes 83.2 16.8 0 Sufficient examples used in teaching 93.7 2.1 4.2 Opportunity to interact/ask question 100.00 0 0 Satisfied with the answers provided 92.6 3.2 4.2

Table 7B: Percentage distribution of item response responding “use of text books” (IPP VII 1997-98) :

Text book Response Non-response Item Yes No Receipt of text book 92.6 7.4 0 Easy to understand 91.6 3.3 5.1 Need to more pictures in the text-book 86.3 6.3 7.4

38

Table 7C: Percentage distribution of item response regarding use of “teaching Aids” (IPP VII 1997-98):

Teaching Aids Response Non-Response Used Not used

Pictures 90.5 4.2 5.3 Charts 86.3 6.3 7.4 Models 22.1 51.6 26.3 Projections 34.7 41.1 24.2 Video films 72.6 16.8 10.5

Table 7d: Percentage distribution of item responses regarding “Use of black board”

(IPP VII 1997-98) Blackboard Response Non-

Response Frequently occasionally Blackboard used 77.9 21.1 1.0

Difficulties in understanding writings in the blackboard

6.3 93.7 0.0

Enough time for taking notes 85.3 14.7 0.0 New and technical words written on board 71.5 27.4 1.1 Drawing of pictures 69.5 30.5 0.0 8. The Socio-economic-demographic and cultural pattern of the female labour

force participation of the North West and The Cape (June’95 –December ’96): Nature of Participation: Principal Investigator. Summary: This research project was undertaken jointly at the Institute of development research, University of North West and at the Institute for child and family development, University of the Western Cape. Because, women are the first educators of the society, because they bring up their own children by imparting their family value system, female employment has shown to affect the qualities of life and of development of the children. This study aimed in understanding the phenomenon of female labour force participation by examining its socio-demographic and cultural patterns and the concomitant problems associated with this process. (Ref: Some Aspects of Complex designs in Survey Sampling by P. Maiti HSRC Publishers, Pretoria, South Africa, pages 37 to 44). The study being exploratory in nature had no explicit hypothesis to test. However, the analysis was done along the following lines.

(i) A woman’s decision and commitment to work outside the home is affected by a

combination of material and non-material factors; (ii) Because of their accumulation of cultural capital stock such as job education

and job training, white women will have higher levels of labour force participation;

(iii) Because of high family responsibilities, married women and mothers will have lower levels of labour force participation than will single women and non-mothers.

39

Findings on the women participation on the basis of the study helped the Government of South Africa take certains measures for maintaining quality of family life in the face of increasing rate of participation of women in the labour force.

Own Activities Associated with the Project: Determination of Sample Size and Its Allocation: Determination of the sample size and its allocation into different strata were based on an extensive analysis of the data on population of the two provinces. Data on urban female population were available from the Census. Distribution of the actual total urban female population into two provinces played a crucial role in determining the corresponding sample sizes for the two provinces. Finally, the required samples were drawn following a stratified multistage design.

The work involved in the project consisted of (a) development of the sampling scheme; (b) determination of the total sample size with its distribution into the provinces; (c) providing training for the people to participate in the data gathering operation; (d) preparing instruction manuals; (e) development of the schedules, of tabulation programme and of the blank

formats for data acquisition and presentation of summarised information; (f) suggesting the form of the estimators for the relevant parameters; Actual

computation procedures for calculation of multipliers and estimators were suggested. The Sampling Design:

The sampling design developed was that of a multistage stratified one with districts, towns, suburbs, households forming the different hierarchical units. The urban female population in the Western Cape Province were stratified into a number of strata, population size being used as the stratification variable. The Questionnaire Design: While designing the schedule, enough care was taken of both current users as well as non-users to obtain information on the basis of matching sample. Exercises were made on the effect of Call-backs on the responses as well as on error analysis due to non-response were made. The Error Analysis:

Table 8A: Interviews Completed in Percentage by the Number of Attempts/Call-Backs No. of Calls 1 2 3 Substitution Proxy Total

Socio-economic Demography

67.32 18.91 2.00 4.901 - 93.13

1 ‘Time of interview’ was not appropriate for some respondents and hence even in spite of 3 Call-backs, some households had to be substituted.

40

Table 8B: Percentage Distribution on Unit Non-Response

Response

Non-response by reasons

Ineligible Contacted, but not

possible to be solicited Solicited,

but refused

Total Socio-economic demographic and cultural patterns

93.13

1.09

5.35

0.43

6.87

9. Survey of family and Community life in the selected communities of the Cape

Peninsula of the Republic of South Africa. (January, 1995-December, 1996). Nature of participation: Principal Investigator. Summary:: This research project was undertaken at the Institute of child and family development, University of Western Cape. On a request from the Institute to the Centre for Statistics (CENSTAT) at the HSRC, I was associated with the Project as the Principal Investigator to help them formulate the problem, develop the sampling scheme and prepare the schedules. (Ref. Some Aspects of Complex designs in Survey Sampling by P. Maiti, HSRC Publishers, Pretoria, South Africa, pages 44 to 47).The study aimed to develop the demographic profiles for the selected communities with a view to understanding the family and the related problems. There were many technical issues on the concepts and definitions of the problem. After several rounds of discussions with the social scientists at the HSRC, the problems were identified and efforts were made to arrive at unambiguous definitions. For example, the meaning of the concept ‘family’ varied greatly from one racial group to the another. This needed standardisation.

The study was to address the critical problem of lack of knowledge about the family and then to develop a data base for systematic analysis of certain family issues, with a view to help many policy makers implement important Government programmes such as Reconstruction Development Programme (RDP) etc. This study became extremely helpful in generating data on many demographic and socio-economic characteristics of the families.

This study helped many policy makers in the government implement many government programmes such as Reconstruction Development programme (RDP) etc. This study become extremely helpful in generating data on many demographic and socio-economic characteristics of the families. Nature of my work: Development of Schedules:

Since most of the questions were of very sensitive in nature, special care was taken in specifying the form and language of the words. Different unstructured forms such as both ‘open-ended’, both ‘closed ended’ or ‘open-closed’ or ‘closed-open’ etc., were used for different items. Since the responses were expected to be arising out of respondent’s belief and faith on different items of information, unstructured questions were used to increase spontaneity in the response of the informants.

The Sampling Design A stratified three-stage design with extensions, clusters, visiting points which formed different hierarchical units was developed and deployed. A non-probability sample was

41

also drawn for understanding (a) structure issues such as public services, their quality and co-ordinations, (b) the acceptability of education, (c) health care, (d) child welfare, (e) employment etc.

The Error Analysis: Table 9A: Interviews Completed in Percentage by Number of Attempts/Call-backs

Number of Calls 1 2 3 Substitution Total Survey of Family and

Community Life 61.32 25.95 3.05 7.19 97.51

Table 9B: Percentage Distribution on Unit Non-Response

Respons Non-Response by Reasons

Ineligible Contacted, but not possible to

be solicited

Solicited, but Refused

Total

Survey of Family and Community

Life

97.51

0.96

1.53

-

2.49

10. Community attitudes and preferences pertaining to cemetery and cremation

related issues in the East Rand in RSA. (May ’95 – December ’96). Nature of Participation: Principal Investigator.

Summary: At the request of the Eastern Gauteng Services Council (EGSC), this research project was undertaken as an academic support programme at the University of the Witwatersrand. As the outreach programme of the Council, I was requested to assist the EGSC as the Principal Investigator in formulation and evaluation of the project. With ever increasing demand of land for burial of the deceased, land was becoming scarce especially in urban areas and also due to an uncertain economic climate, the problem to find a suitable land for cemetery development was becoming a matter of great concern.

An alternative to burial in certain communities is the cremation, which unlike to burial, is less land demanding and probably economic also in the long run.

Before the council proceeded to undertake the task of building a public crematorium in the western urban side of Eastern Gauteng, it was therefore necessary to find not only the attitude of the people, but also to get an estimate of the percentage of the people from the community accepting it. (Ref: Some Aspects of Complex designs in Survey Sampling by P. Maiti; HSRS Publishers, Pretoria, South Africa, Pages 54 to 63). The study helped the policy makers find not only attitude of the people, but also get an estimate of the percentage of people accept the concept of building a public cremation in the Western urban side of Eastern Gauteng. Nature of my Work: Combined use of a probability and a non-probability sample: It has been observed on many occasions, that a probability sample alone can not help in drawing valid conclusions on some issues. In this case too, a need was felt on the realisation of a non-probability sample from community based social groups of burial

42

societies, church groups, old generation CBO’s as well as new generation CBOs, and of government-officials. Hence both a probability sample and a non-probability sample were drawn to obtain relevant data. Sampling Design:

The probability sample was realised through a stratified multistage design. The sampling design developed was that of two stage,one with suburbs as first stage units and enumeration blocks forming the second stage units. Different racial groups formed the natural strata for allocation of a total of 480 households – firstly into different racial groups, and subsequently allocations to within particular group were made into different areas. Use of aerial photographs were made for selecting the ultimate units.

Extensive analysis of the available data on population of different racial groups was made to allocate the total sample size first into different racial groups and then for a particular group into different areas. In some situations, strict proportionality was not possible to be maintained, as the sample size in those cases turned out to be too small to be considered.

Questionnaire Design: Three categories of questionnaires of the open-closed-type were developed and used for collecting data.

(a) household schedules (a probability sample); (b) moderate schedules from community based social groups selected purposively (a

non-probability sample) and (c) another set of moderate schedules for key informants such as leaders in the Transval

Legislative Council (TLC), Provincial Government, the EGSC, and prominent religious leaders etc. selected purposively (another non-probability sample).

Exercises were made to examine the effect of Call-backs over the response and also an error due to non-response was analyzed. The Error Analysis: Table 10A: Interviews Completed in Percentages by Number of Attempts/Call-backs

Number of Call-backs 1 2 3 Substitution Proxy Total Community Attitudes 69.05 20.52 6.15 - - 95.72

Table 10B: Percentage Distribution on Unit Non-Response

Respons Non-Response by Reasons Ineligible Non-

Contacted Solicited,

but Refused Unable to Answer

Others Total

Community Attitude

95.72 0.79 0.89 0.96

0.64

1.00 4.28

43

11. Identification of Other Backward Class (OBC) in West Bengal (1994): Nature of Participation: Member Coordinator. Summary: At the request of the backward commission for West Bengal, this research project was undertaken at the Indian Statistical Institute, Kolkata to provide an independent evaluation on the class of different communities. The evaluation was made on the basis of a household survey with the following major criterion.

(i) The degree of economic affluence of the specified communities; (ii) The extent of participation in different levels of education from the above

communities, and (iii)The extent of involvement in different respected professions from such

communities. The study helped the Government of West Bengal identify who could be included in the other backward classes. This eliminated the confusion on some of the commodities, whether they should be considered as belonging to other backward class. On Sampling Design: A three stage stratified sampling design with subdivisions as the strata was adopted. Within each stratum, different hierarchical stages were as follows:

(i) Police station, as the first stage unit; (ii) Village as the second stage, and (iii)Households as the ultimate or final stage units.

At each stage of selection, the required number of sampling units at that stage was drawn by simple random sampling. Estimation Procedure: The sample having been drawn in a probabilities manner, the estimates of the population characteristics were obtained through appropriate statistical formulae. Consider one subdivision (stratum); let

K = no. of police stations in the subdivision, and k= no. of police stations in the sample, drawn by SRSWOR.

Next, suppose the no. of villages in these sample police stations are

kNNN ,,, 21 and we have selected

knnn ,,, 21 villages, from these police stations by SRSWOR.

44

Finally, suppose that from the thj sample village, of thi police station, we have taken ijn'

households from ijn households of the thl caste/community (l = 1, 2, …,5), then the

multiplier or probability weight of any sample household of the thl caste/community of the thj sample village of the thi p.s. is

ij

ij

i

iij n

nnN

kKM

'..= .

Such multipliers were computed for all sample households drawn from all the strata. Now, consider any characteristic of the ),,( ji th sample household, like number of females participating in a gainful work or number of matriculates in the household. Denote any such characteristic of the household ijy

then

ijiji j

yMY ∑∑∑=ˆ

gives the estimate of the total of y in the stratum and adding over strata, we get estimates for districts or for all districts combined. 12. Evaluation of Total literacy campaign (TLC) in West Bengal (1990-91): Nature of Participation: Member Secretary. Summary: The objective of internal evaluation was two fold: a) evaluation of literacy attainment of the learners of the district in terms of the norms of National literacy Mission (NLM) and b) general observations on other aspects of TLC in the district (such as motivational, organizational, technical etc.,). On the basis of the study, Government of India was able to declare a district as a literate one and helped the government understand, if her objective of achieving total literacy has been ful filled or not. On Sampling Design: The internal evaluation of ‘Total Literacy Programme’ sought to assess the current state of attainment of literacy among the learners. This was done principally through representative, scientifically designed and conducted sample investigation of the literacy status. On consideration of an efficient sampling design, a two stage stratified cluster sampling design was adopted; the four subdivisions of the district formed the four rural strata and similarly, nine municipalities constituted nine strata for urban areas. The Gram Panchayat (GP)S were considered as first stage units and the learning centers within the GPS formed the second stage or ultimate stage units. At both the stages, Simple Random Sampling was adopted.

45

Estimation Procedure. In major cases, the design was self weighting. In the remaining cases, disproportionate sampling was introduced to compensate for differences in sample rates. These weights were mostly based on the original probabilities of selection. This is evident from the following formulations. Let d = Subdivision (d=1,2,3,4; 1 = Barasat, 2 = Barrackpore, 3 = Basirhat, 4=Barrackpore) t= Sector (t=1,2; 1=Rural, 2 = Urban) h= GP (h = 1,2; dG ; dG being the number of GPS in subdivision d) k=l.c. (learning centre) (k=1,2,… hL ; hL , the number of learning centers in thh GP) S= Six (S=1,2; 1 = Male, 2 = Female) p= a learner (p=1,2,… s

kn ; s=1,2, k=1,2,…, hL ) Let

dthkpy stand for the value of the characteristic y associated with thp learner in thk learning

centre from thh GP in the sector t of the subdivision d.

Defining =

=)and,,(givenotherwise,0

)and,,given(1for1)(

khtdkhtds

Sy dthkp

One can have

.nsubdivisioawithinsectortheofG.P.

tobelongingcentrelearningfromlearnersmaleofNumber)(1 1

dt

hKSyi

S

ththdthkp

n

p

sk

∑∑= =

=

Similarly, the number of female learners from the above classification can be obtained, obviously,

)(1

Syn dthkp

p

z

sk ∑∑

=

= ; ( )(Sydthkp =1, for both s = 1, and 2)

denote the total number of learners in the thk learning centre from the thh G.P. in the sector t within a subdivision d.

46

Let

hl = number of learning centres sampled from thh G.P.;

hL = total number of learning centres in the thh G.P. Now the estimate of the total number of learners in a selected (G.P,t) classification

∑∑∑===

)(111

sylL dt

hkp

n

p

z

s

l

kh

hskh

Writing,

)(111

)( syn dthkp

n

p

z

s

l

k

tdh

skh

∑∑∑===

,

We have an estimate of total number of learners for the sector t in a subdivision d as

)(1

tnlL

gG d

hh

hg

hd

dd

∑=

and finally the estimates of the total number of learner for the sector t, in all the subdivisions was found to be

)(1

4

1tn

lL

gG d

hh

hg

hd

d

d

d

∑∑==

Putting t = 1,2, estimates of total number of learners were obtained for rural and urban at different levels of aggregation. 13. ISI-ONGC Collaborative Research Project (1985-1992) Nature of Participation: Member Statistician: Summary: At the request of Oil and Natural Gas Commission (ONGC) of India, the Indian Statistical Institute evaluated the economic and physical consequences of various strategies for action in different basins in India. Both economic and stochastic models were used for estimating reserve and the per unit cost of hydrocarbon. (Ref. Estimation of Discovery and Production costs of Hydrocarbon with some application to Indian Data, Indian Statistical Institute, Calcutta). Transformation of the real life problem into the statistical one required a series of discussions with the technical experts working at the different levels of the organization. In fact, the project formulation was not a routine work: instead, definitions and other related concepts were defined, developed and redefined into the frame work of the

47

present problem. For example, the ‘reserve in place’ was distinguished from the ‘recoverable reserve’. On the basis of the findings on discovery of hydrocarbon as well as production cost of oil and gas, Government of India was able to revise per unit of oil and gas prices.

The following were the technical novelties among others:

(a) The available primary cost data were in the form of well-wise cost. The well-wise cost figures were measured at current prices, and therefore, the cost data for different years were non-comparable. In order to make them comparable, it was necessary to deflate them using a suitable index number of well-drilling cost. No such index number was available which could be used for the purpose of deflation of cost figures measured at current prices and for this, a new index was constructed and applied to the given data;

(b) The question of how to aggregate and what economic models to choose had to be resolved;

(c) To test on the constancy of success ratio in hydrocarbon exploration, data were examined through a number of statistical devices some of which were based on graphical representations, while the other explored standard statistical techniques.

(d) A fully Bayesian Hierarchical method which provides better estimates for errors in estimation and prediction was sought for. But because of analytical complications, an empirical Bayesian view was taken in predicting the (n+I)th discovery, given the data on past discoveries. Two types of simulation estimates were provided (i) one based on the assumption of an approximate Gamma distribution of the field sizes where as (ii) the second alternative used a Gamma population and employed the classical method of “importance sampling” for adjustment. Both methods involved novel methods of simulation developed by us.

14. Domestic Tourists Survey in Orissa (1988-89): Nature of Participation: Member Coordinator Summary: One of the major objectives was to determine the factors that could be ascribed to the promotion of tourism of India. Apart from determining the extent of revenue the country could earn, this study also aimed to investigate certain sociological problems related to the tourism industry. Along with the development of methodology, it became necessary also to observe.

(i) Socio-economic distribution of tourists in India; (ii) Regional distribution of tourists across the country; (iii)The existing infrastructure facility available in terms of accommodation,

transport, medical help etc., (iv) Average cost incurred by a tourist during his/her travels for travelling, boarding

and lodging etc. (v) Influx of tourists during a period at a particular tourist spot.

48

The need of developing such infrastructure facilities as different types of places of accommodation, connectivity to road, rail and of general administration was felt by the Government of India on the basis of the findings of the study. It suggested the government on the type of hotels affordable to the middle, and upper class of people, as it was found that the major contributors to the flow of tourists were the middle class and/or upper middle class people. Tourists were distributed in different towns and within a town, they stayed in different types of dwellings as 1: Hotels; 2= Dharmasalas; 3 = Circuit House; 4=Irrigation Bungalows; 5 = Forest Department Bungalows; 6= Railway Bungalows; 7=Youth Hostels; 8=Open air. . With a view to the proper representation of all types of categories, an appropriate scheme was designed and adopted for drawing samples of tourists. Some technical novelties of the work were as follows.

(i) Among other objectives enquiries were also directed towards finding availability of existing infrastructure facilities in terms of accommodation, transport (road, rail, air), medicine etc. This required redefinition of a tourist. Normally, a tourist by definition is a person who visits places of historical monuments, pilgrimages etc. According to the objective of the study, any person for any reason whatsoever, requiring accommodation to spend at least one night should be considered as a tourist and hence because a member of the target population. Therefore, the usual definition of a tourist became unusable and was defined according to the objective of the study. Otherwise, target population considered could have been under coverage.

(ii) For a stratified random variable, choice of the stratification variable must be correlated with the study variable; that the administrative zones should not always form the different strata has been examined in this case; Degree of concentration has appeared here to be the stratification variables.

(iii)A multistage sampling design was developed and deployed. It may be interesting

to note that and was stratified into six strata

From each stratum, a probability sample of a period of 10 days was selected for visiting these time points. Towns/cities formed the second stage units, while dwelling/accommodation places within the selected town and the tourists within a particular type of dwelling place served as the third and fourth stage units respectively. At all the stages except at the final stage, simple random sampling was introduced in selecting the pre-determined number of units from that stage.

49

At the final stage, a random sequence like D,S,S,M,M was generated and each investigator was asked to follow strictly the random sequence thus generated for selecting the room of a particular type in each selected hotel, where D = double room, S = single room, M = multiple (Dormitory). 15. Costs and Benefits of Rural Electrification REC Project (1975-76): Nature of participation: Project Assistant. Summary: Primary objectives of the study were (i) quantification, as far as possible, of all relevant costs and benefits of the selected REC project with reference not only to the time, when the study was undertaken and when the projects were only in their second or third years of life, but also to the future costs and benefits for each of the remaining years of the life span of the projects (ii) identification of factors that facilitate or obstruct full realization of benefits flowing from electrification in rural areas. It because easy for the Government to decide, if the existing projects under Rural Electric Corporation could be extended further on the basis of cost-benefit analysis of rural electrification or not. Subsidiary objectives:

(i) Assessment of share of benefits enjoyed by different socio-economic groups in terms of proportions or percentages;

(ii) Assessment of socio-economic impact, for example, changes in employment, forming techniques and migration to urban centers in terms of rates, proportions etc.

Some Technical novelties of the work: On Sampling Design: Household /enterprises were sampled according to two-stage sampling scheme, using probability samples. In the first stage all the villages were listed and certain particular relating to use of electricity were tabulated. This enabled stratification of the villages into a number of homogenous groups. At the second stage, from the elected villages, household/enterprises were again classified depending on the type of use of electricity. On stratification: On examination of the village wise lists of connections, a definite structure was found to be present in the distribution of connections by villages. A small number of villages were intensely electrified and accounted for approximately half of all connections of each type.

50

With this pattern in distribution, a random sample often, twenty or even 50 percent villages is unlikely to yield a reasonable number of connections of all types. Moreover, the intensely electrified villages were to display conditions which favoured utilization of electricity. It was therefore deemed essential that the villages be grouped into a number of strata according to the intensity and type of use of electricity. For the second stage of sampling, an unorthodox schemes was adopted. Instead of selecting the households/enterprises from each of the villages selected in the first stage, all such households/enterprises were collated together. From this single list of second stage units, the required number of samples was drawn at random. This sampling was obviously less efficient from the point of view of sampling error, but was necessary to ensure adequate representation. 16. Calcutta Urban Poverty Survey (1976): Nature of participation: Project Assistant. Summary: The objective of the survey was to study the nature, extent and the causes of poverty of the segment of population living in Calcutta pavements, which might be different from rural poverty, since there is a sharp difference in the livelihoods of rural and urban people. The important features identified for investigations were

(i) Type of living unit; (ii) Employment; (iii)Migration,

and (iv) Education.

Other than the above important features, several more were studied viz., type of shelter, type of toilet facilities, per capita weekly income etc., as certain aspects revealing standards of living. The study helped the government agency understand the nature, extent and the causes of poverty of the pavement dwellers. It also determined distribution of pavement dwellers by socio-economic class, by their origin of domicile, from where they migrated to the city and residing in the pavements. This helped the government prepare the plans for improving the rural economy to resist migration towards the city from the rural sector of the neighbouring districts.

(i) Coverage:

The survey extended over the whole of Calcutta Corporation area excepting 9 inaccessible blocks.

51

(ii) Sampling Design:

The broad design was a two-stage one with blocks as first stage units and pavement dwellers’ households as second stage units.

(iii)Stratification: Phase I: The 5127 NSSO (National Sample Survey Organisation of Government of India) blocks were visited and the approximate number of pavement dwellers residing in each was obtained through local enquiries.

These blocks were then grouped into 8 zones according to their geographical locations. Then the blocks were further sub-divided into 4 strata according to the approximate number of pavement dwellers in each. Thus the 5127 blocks were stratified into 32 strata of 8 geographical zones and 4 size classes of the “number of pavement dwellers.

Phase II: The blocks were grouped into 4 geographical zones, combining two contiguous zones of phase I. Thus 5127 blocks were stratified into 16 strata of 4 geographical zones and 4 size classes.

Sample Selection:

Phase I: 64 blocks were selected by SRSWOR, two from each stratum. These sample blocks were divided into two half samples. Phase II: 32 blocks were selected by SRSWOR, two from each stratum. Next 8 more blocks were sub sampled from the 64 samples blocks of phase I.

Estimation Procedure: The following notations were used for obtaining various estimates.

r: Subscript for thr zone; S: Subscript for ths stratum ; i: Subscript for thi surveyed block; j: Subscript for thj household; m: number of surveyed blocks; h: number of households ; M: total number of blocks ; y: value of a characteristic under study; Y: total of y for all the subsamples combined; Y : estimates of the total of y for all the sub-samples combined.

52

Estimate of Total An estimate for the total of y is given

∑ ∑∑∑=

=r

Mrs

mrsrsij

hrsi

jsy

mrsMrsY

1

ˆ

mrsMrs is known as the multiplier.

Pooling of phase wise estimates:

21

2211

ˆˆˆnn

YnYnY++

= , in is the number of blocks surveyed in the thi phase (i =1, 2)

17. Health and Socio-Economic Survey of Calcutta Metropolitan Development Area (1976-77); Conducted by Metropolitan Development Authority in Collaboration with the Department of Health, Government of West Bengal. Nature of Participation: Project Assistant. Objective: Government of West Bengal entrusted Calcutta Metropolitan Development Authority with responsibility of improving medical facilities in the Calcutta Metropolitan Development area. For this, it became necessary for them to know what facilities were available, what more would be required and where they should be located. The Government and CMDA thought, it was wise to carry out a fact finding study of the available medical facilities in the area. The Indian Statistical Institute was requested to obtain a survey based estimates of the stock of the existing medical facilities and their functioning in the Calcutta Metropolitan Development Area. This also aimed to assess the health condition of the population living there in with reference to their social and economic conditions with a view to estimate the demand on medical service of the different disciplines. Sampling Design: The sampling design used was that of a stratified simple random sampling without replacement of medical care institutions. A stratified multistage sampling was also adopted for drawing a sample of households. The study helped policy Decisions Committee to improve upon the medical facilities in the Calcutta Metropolitan Development Area. On the basis of the findings, the policy makers were able to know what medical facilities were available, what more were required and where they should be located in Calcutta Metropolitan Development area. This helped the government in extending medical facilities in the area.

53

19. Important Professional Work:

1. Worked as a chair person of the sessions of the two – day UGC sponsored national seminar on decentralised planning, Department of Economics, Vidyasagar University, March 20 – 21, 2003.

2. Worked as a chair person in the panel discussion on information support system, Golden Jubilee Conference of Indian Association of special libraries and Information Centre (IASLIC), December 31st 2004.

3. Worked as a member of the advisory committee for the two – day national conference on “Recent Trends in Estimation and Optimization: Theory and Applications” to celebrate platinum jubilee of the Institute of Science, Nagpur, India, to be held during January 1 to January 2, 2006.

4. Worked as a member of the Advisory Committee on Geological Information System (GIS) to assist in (a) developing the method by which data could be generalised on a geographical basis, (b) developing sampling frames and drawing samples, (c) upgrading the human resources at the CENSTAT HSRC, Pretoria, South Africa with new techniques of sampling and to take part (d) in different research programmes for the centre of statistics.

5. Worked as a member of the technical committee set up for the project “Human Resources Survey of Black Professionals in South Africa”. The followings were the responsibilities of the technical committee.

- To design the scheme and study according to the guidelines and definitions set up by the steering committee;

- To train the supervisers and field workers to collect data from the respondents;

- To provide all the technical details in connection with the project including time scheduling, questionnaire design, sampling design and weighting the sample data for calculation of the estimates.

6. Dr. P. Maiti and the Pslam Programme: (An extract from the minutes of September 12, 1995 CENSTAT, HSRC/RGN,

PRETORIA, South Africa) “….. Dr. Maiti will contact Professot Stoker to try and get the exact sampling design --- where upon he will develop formulae to test the data……… The Computer Centre will write the Pslam programme……” The project was executed according to the above guideline. Exercises were carried out to examine, (a) if different response rates existed in various categories/subgroups which were formed with respect to (i) certain demographic variables such as geographical location, degrees of urbanisation and population group and/or (ii) with respect to certain biographical variables such as sex, age-group, marital status, occupational group and level of education. (b) if these varying proportions resulted in skewness in the realised sample in respect of such variables. This was done by comparing the sample data with published statistics from census. The procedure followed was that of calculation

54

of proportion of units belonging to the different subgroups from the available census data and then multiplying them by the sample size n, to obtain a set of frequencies which were compared with the frequencies of the sample data. No significant deviations between the sample data and census data were detected. While calculating the weights, the factors of unequal probability sampling, non-response and variation of the varying weights were taken into account.

7. Conducted an workshop on Sampling (April ’95 – December ’95): As a specialist consultant in Survey Methodology at the Centre of Statistics of the Human Sciences Research Council, my expertise was rendered to HSRC statisticians, researchers and other stakeholders in investigating their various theoretical and practical problems on sampling. Workshops on ‘Sampling Theory and Practice’ were conducted at the centre for statistics every two weeks during the whole of 1995, according to the following time table. April: 19/4/95; May 3/5/95 and 17/5/95; June 14/6/95 and 28/6/95, July: 12/7/95; August 2/8/95 and 30/8/95; September 13/9/95 and 27/9/95. October: 11/10/95 and 25/10/95, November 8/11/95 and 12/11/95; December 6/12/95.

55

20. Teaching Activities: Teaching: (1995 - )

Courses Subjects: Number of Lectures Theory Practical

Centre of Statistics, Human Sciences Rresearch Council, Pretoria, South Africa 1: Lecture Series on Survey Methodology and Basic Principles of Sampling: March 25-29, 1996 2: Lecture Series on Sampling Methodology: April 14-May 15, 1996 3: Lecture Series on Advanced Sampling Techniques: August 14-18, 1996

Sampling

Sampling

Sampling

- -

Indian Statistical Institute, Kolkata 1: Master of Science in Quantitative Economics(MSQE)

(1997-1988), (1998-1999), (1999-2000), (2000-2001), (2001-2002), (2003-2004), (2004 – 2005), (2005 – 2006),(2006-2007) (2007-2008),(2008-2009)

Statistics

50

20

2.: B. Stat. (1996-1997), (1997-1998)

Economic Statistics and Official statistics

34

3.: PG Department of Statistics The MEDUNSA, Pretoria, South Africa (1996-1997)

Statistical Quality Control

Full one academic

year

4.:INTERNATIONAL STATISTICAL EDUCATION CENTRE: (ISI).

(1998-1999), (1999-2000), (2000-2001), (2001-2002), (2002-2003), (2003-2004), (2004-2005), (2005-2006) (2006-2007), (2007-2008) (2008-2009), (2009-)

(a) Sampling (b) Economic Statistics II (c) Economic Statistics III (d) Sampling at the

Specialisation Course under UNDP

(e) Sampling at the Specialisation course in the regular course

(f) Probability and statistical methods in the regular course

30 35 35 45

40

20

15 15 15 15

--

--

5: Tenth Course/Workshop on Sampling Design for Household and Establishment Surveys (organized jointly by Central Statistical

Sample Size Determination in Survey Sampling (Prepared a training handbook for this purpose)

56

Organization, Ministry of Statistics and Programme Implementation, Govt. of India and United Nations Statistical Institute for Asia and the Pacific (UNSIAP), Japan): 18 Oct. – 12 Nov. 2004, ISI, Kolkata 6: The ISI sponsored three day workshop on Demography and Population Dynamics with emphasis on Survey Sampling Methodologies at the Department of Mathematical Sciences, Tejpur University: 26 Oct.-29Oct.2006.

Nonsamplig Errors-classification, And quantification (Prepared a training handbook for this purpose)

7:UGC-Sponsored Refresher Course on “Decentralisation, Planning and Participatory Development” organized by Dept. of Economics with Rural Development, Vidyasagar University: Dec. 12, 2005 – Jan. 06, 2006.

Development of Statistical Information System (Prepared a training handbook for this purpose)

8: UGC-Sponsored Refresher Course in Statistics organized by UGC Academic Staff College, Sambalpur University: Dec. 04-24, 2008.

Sampling Methodologies (Prepared a training handbook for this purpose)

57

21. Authored Text Books: About the Books

Some Aspects of Complex Design in Survey Sampling HSRC / RGN Publishers; ISBN 0 –7969 – 1857

Pulakesh Maiti

Abstract The two major features of any survey design are the sample design and the schedule/questionnaire design. Only occasionally, we do come across a book that deals with different complex designs when applied to real life situations. Many difficulties, especially those that arise in actual field conditions, make it almost impossible to use standard results and hence a compromise between the theory and Practice is very much needed. The purpose of this write up is to show in a practical manner how to arrive at a sample design for a complex survey after giving due considerations to practical constraints such as time and cost and to develop a schedule or a questionnaire in an efficient way. The approach is essentially practical. The material therefore emphasizes on (i) how a complex survey design is developed, (ii) how one can compute multiplier from each record / piece of information to build up estimator, and how one can make an assessment of the reliability of the estimators in a multistage design with minimum effort. Extensive use has been made of some of the projects, undertaken at the Indian Statistical Institute, Kolkata and at the Human Sciences Research Council, Pretoria, South Africa. Since I believe that a survey practitioner needs to be exposed to basic principles of sampling, there follows a brief discussion, not so much on providing proofs, but on understanding why results are true, what makes these work, and how they are applied. It is hoped that researchers in various different disciplines may find this useful when designing surveys for their own research problem. This write up is intended to help non-specialists in sampling in their own work.

FOREWORD Dr. Pulakesh Maiti has written a fine monograph on the theory, practice and historical evolution of survey sampling. He has drawn on existing theory, including his own work, as well as his experience and expertise gained through practical work at the Indian Statistical Institute on such different aspects of surveys as design, collection, scrutiny, editing and analysis of data. Altogether he presents a very balanced view of the subject that includes some of the latest methodologies. Being a Bayesian, I myself would have liked to see a bit more of the Bayesian approach to these problems. But it is true that most practitioners of survey sampling are still not Bayesians though they are now more sympathetic to Bayesian ideas than before. I was also very pleased to see many references to surveys done in South Africa. This is a indeed very appropriate because this way Dr. Maiti has validated the basic principles of survey sampling over a bigger set of applications and drawn attention to important country specific problems. I am sure Dr. Maiti’s application-oriented but scholarly approach to his subject will gain him many readers with very diverse backgrounds. I like to think of this monograph as indicative of great things that can happen if developing countries share resources and participate in innovative activities of mutual concern. Professor J. K. Ghosh Jawaharlal Nehru Professor, Indian Statistical Institute President, International Statistical Institute (1993 – 1995)

58

Sampling and Estimation Procedures for Inverse Multinomial Sampling Associated with Single, Multiple and Joint Events.

H. S. Steyn and P. Maiti

HSRC / RGN Publishers; ISBN 0 – 7969 – 1822 – 8.

Summary In traditional sampling surveys the total size of a sample to be drawn from a statistical populations is normally considered as a given parameter. However, the situation changes when the focus is on adequate representation of a specific subgroup in a population. It is often necessary to introduce different kinds of randomisation procedures to ensure adequate representation of a specified group among the subgroups of a population. This type of target sampling is known in statistical literature as inverse sampling. The inverse sampling techniques introduced and used in this monograph ensure a representative sample containing an adequate predetermined number of units possessing a specified character or belonging to a subgroup of special importance within the population. It briefly introduces the existing theory on inverse sampling and these expands the theory to include multiple events (i.e., events that can occur more that once per trial) as well as joint events. An effort has been made to keep the mathematical statistical procedures contained with text on a level that will be accessible to researchers in the Human Sciences with a limited background in mathematical statistics. Some mathematical derivations are included as appendicies. Some ideas on sampling from labelled population units are mentioned briefly, but because of their more advanced mathematical nature, be dealt in a separate publication. 22. Book-Chapters:

1. Interactive Linear Models in Survey Sampling (2014): Essays on Mathematics and Statistics, Edited by Codueta Stoica, Athens Institute of Education and Research (ATINFR) pp. 141 – 145. Considered is a liner ‘interactive’ model in the context of survey sampling. The situation arises when investigator and/or supervisor interventions are contemplated in the responses. An appropriate linear model is introduced to represent the response profiles(s) arising out of each respondent-cum-supervisor combination as per the panned ‘design lay out’. Two situations (blinded and unblended submission of responses) are differentiated and corresponding data analysis techniques are discussed. Variance components are assumed to be known in the study.

2. Indian Statistical Institute – Numbers and Beyond (1931-1947) (with J. K. Ghosh and Anil Bera) (2007): To appear in Project of History of Indian Science, Philosophy and Culture (PHISPC) Volume 1, “Science and Modern

59

India: An Institutional History C ‘1784-1947’ edited by Professor Uma Dasgupta. The project to study the history of Indian Science, Phylosophy and Culture from the earliest times to 1947, funded by the Ministry of Human Resource Development, Government of India, has been taken up with the objective to launch a comprehensive and interdisciplinary study of scientific, philosophical and cultural heritage of the Indian civilization from the past to the present times.

The volume on “Science and Modern India: An Institutional History C ‘1784-1947” includes our article‘Indian Statistical Institute – Numbers and Beyond (1931-1947)’ in its second section, which has been made up of chapter histories on the major science institutes that were founded in those years which were contributing silently or overtly to the nation building effort.

The present essay is essentially in two parts, one on Mahalanobis (the early years), the other on both Mahalanobis and Indian Statistical Institute (ISI) during the period under review. Part 3 suppplements parts 1 and 2 and adds details on rthe first two decades of the ISI as well as some earlier and later events.

Part 1: Prasanta Chandra Mahalanobis – the Early Years (1897 – 1931). Part 2: The Statistical Laboratory and the ISI. Part 3: Supplements and the Appendices.

3. Development of Statistical Information System (SIS) for Decentralized Planning (2004), pp. 198-248: Participatory Decentralized Planning in India: Issues of Finance and Statistics Information,Edited by Professor Sau , FIRMA KLM Publication, Calcutta. The Government of India had requested Asian Development Bank for technical assistance to strengthen local Government of Madhya Pradesh. The then Price Water House India (PWI)-presently Price Water House Cooper was one of the prime contractors to the Bank for this purpose. An inter-firm agreement between Price Water House India and the Indian Statistical Institute was made to assist for data requirement, data procurement needed for planning and efficient decision making by local bodies. The Statistical Information System (SIS) developed at the Indian Statistical Institute at Kolkata as a part of the assignment entrusted to the Institute by PWI is presented here. The SIS has been developed in accordance with the 73rd and the 74th Amendments of the Constitution of India.

23. Summaries of Major important professional activity: Under Major Important Professional Activity :

1. Serving as a member of the working group of 73rd round of NSS (June 2015 – June 16) On being inducted as a sampling expert in the working group of 73rd round of NSS on an invitation from National Statistical Commission, expertise on the development of sampling design, designing the enterprise schedule and in the selection of first stage units in rural and urban areas was provided. For the first

60

time, it was suggested that Census Enumeration Block (EB) should be taken as the first stage units in urban areas.

2. Serving on the working group of Drug Abuse Survey (November 18, 2014 – April 10, 2015) Knowhow on the development of sampling design was provided. The sampling design was a dual approach i.e., household based survey for commonly used (lied) drugs and an alternative survey method known as Respondent Driven Strategy (RDS) technique to capture illicit drugs. For this, licit and illicit drugs were identified. Clarifications on the concept of ‘Drug Abuse’ and ‘Drug use’ were sought for. Suggestions were offered on tackling the issue of multiplicity of respondents across household based survey and RDS technique. A new estimation procedure in case of the frame having multiplicity problem was suggested.

3. Serving as a member of the working group of 69th round of NSS (June 2012 – December 2012 ) The National Statistical Commission (NSC) invited me to serve as the sampling expert in laying down sampling design, methodology and strategies to reduce non sampling errors for 69th round of NSS. The National Statistical Commission (NSC) was set up by an Apex body by an Act of Parliament Government of Imdia and empowered in monitoring statistical activity in the Country. The following contributions have been made towards (i) Estimation of number of slums: A sampling methodology has been

developed in estimating number of slums in the face of having no frame for the slums. The sampling methodology thus developed suggests the number of slums can be estimated unbiasedly from the survey with the help of an additional information on the number of blocks linked to a slum of which the sample blocks is found to be a part. If there are more than one slums intersecting a sample block, then additional information needs to be collected in respect of each of the slums. Serving as the Chairman of the sub-committee on

(ii) Imputation of rental values: The back ground is the issue of compilation of house rent component in the consumer Price Index for rural areas on similar lines as is being done in respect of CPI for urban areas. The issue was discussed in the Technical Advisory Committee on statistics of Price and cost of living Index. The TAC recommended to refer the matter to NSC for exploring a possible methodology. Accordingly, the issue was placed before NSC. The recommendation of the NSC in this regard was as follows.

61

“With regard to the reference made by the Central Statistical Organisation (CSO) (NAD-PCL unit) on the issue of imputing house rent in respect of owner occupied houses in rural areas, it was felt that relevant data needs to be collected by the NSSO for the purpose of including weights on the item in the price collection module by the CSO. It was decided that the matter be referred for consideration in detail by the 69th round working group of NSS”.

As recommended by NSC, the working group of 69th round of NSS may therefore consider the above issue and explore a simple alternative methodology for imputation of rent in respect of self-owned houses in rural areas so that the same may be adopted in the next quinquennial round of consumer expenditure survey by field staff of NSSO. As subcommittee was formed and the following documents were prepared. (i) Development of a methodology/number of methodologies for imputing

rental values was/were made; Different imputation methods were discussed;

(ii) A document with underlying mathematical justification was prepared and the methodology proposed there in was used to obtain an unbiased estimate of the number of slums in the face of having no frame for the slums;

(iii) A document on the estimation of coverage error was prepared; (iv) A document dealing with the problem of non-response was prepared. This

aimed in estimating the response probabilities. (v) A document has been prepared in identifying causes of non-sampling

errors in the NSS-field work. On the Recommendation on Selection of Urban Frame Survey UFS-blocks:

As per earlier guidelines, a norm of about 600-800 by population size was used for formation of UFS-blocks. Since the blocks were supposed to be of more or less of same size, it had been the convention for the NSS to select sample blocks by SRS.

In 2007-2012 Urban Frame Survey (UFS), it was not possible to maintain said norm strictly at the time of formation/ updating of UFS blocks. A study of 61st round sample blocks reveals that as high as 28% of the sample blocks had a population of size 400 or less. On the other hand about 14% of sample blocks had a population size of 1000 or more. In other words, the UFS blocks, the way they were being formed, do vary substantially in size.

62

Therefore, since under 2007-2012 UF Survey, the UFS blocks would vary in size, it has been recommended that the UFS blocks for NSS surveys should be selected by PPSWR and not by SRSWOR, as it is the present practice.

Field visits: Field visits were made on the sample basis by myself according to the following schedule; This helped understand the causes of different kinds of non-sampling errors. A document has been prepared in this direction.

NSS-field visits August 2012 – December 2012: Field visits were made in the following states according to the following schedule.

State District Village/town/city Duration of field visits

Tripura West-Tripura Braspur Badharghat

(IV unit 0001)

August 21, 2012 – August 24, 2012

West Bengal Darjeeling Pumong tea garden Siliguri MC

(IV unit 0003)

September 11, 2012 – Sept. 14, 2012

Jammu and Kashmir

Jammu Jandyal Jammu (IV unit 0039)

September 25, 2012 – Sept. 28, 2012

Delhi North West Delhi MC (IV unit 0038)

October 29, 2012 – October 31, 2012

Gujarat Vadodara Dunad Vadodara MC (IV unit 0086)

November 20, 2012 – Nov. 23, 2012

Andaman and Nicobar Islands

Carnicobar South Andaman

Malacca Port Blair

(IV unit 0005)

December 04, 2012 to Dec. 07, 2012

West Bengal North 24 Parg. Kolkata

Chandpara Kolkata MC

December 17, 2012 to Dec. 17, 2012

4. Working as an Analyst (2011-12): Appointed as the Analyst of village Resource Mapping by the Directorate of Micro & Small Scale Enterprises, Government of West Bengal, India, and rendered my expertise to guide the Director at regarding the fields to be considered for identifying potential resources in specific areas of different districts which might ultimately facilitate industrialization in the state.

63

5. Working as a collaborative Scientist for some of the Projects undertaken by the Department of Economics and Statistics (DES), Tata Services Limited (May 17, 2010 – August 16, 2010):

Projection of Demand of thirty two items: Methodology on the determination of the projected market size for consumption of items for the determined period 2015 and 2020 have been suggested. This has involved estimation of the Engle Curve/Demand function of each of the thirty two items and that of the total expenditure based on the data on rural consumption for the year 2004-2005. Before actual fitting distributions, few tests on graphical representation were carried out. However the distributions fitted to the given data on different items of consumption appeared to follow long normal distribution. Projection case 1: 2σ , which indicates the measure of inequality remains constant and the elasticity is also of constant type. The Engle Curve/demand function is of the form

( ) baxxyE = , y and x being items wise and total per capita expenditure. It was shown that,

)(0

0µµ −= tbt eyy

Where

ty = Average demand at the future time point ‘t’

0y = Average demand at the present time point ‘0’

tµ = Mean of the distribution of x, at time point ‘t’

0µ = Mean of the distribution at present time point ‘0’ With the given data, demands for thirty two items, some of which were luxary and necessary type and some of which were of inferior goods were projected at a future time point. Case 2: 2σ undergoes changes. From the relation

σ−= FF tt1

64

Where,

)(1 xF = Proportion of aggregate income through all those earners whose income is x≤ units. F(x) = Proportion of earners having income x≤ .

1Ft and Ft are the points abscissa of the normal probability curve up to the areas 1F and F

respectively. The above relation has helped in answering the querries of the following type. Suppose “top 10% has 30% of the income in 2010” changes to “to 10% has 35% of the income in 2015”, what happens to 2σ .

6. Providing Technical Assistance for Piloting Below Poverty Line (BPL)

Census (May 2010 – 2011):

At the request from Principal Secretary to the Government of West Bengal, Panchayat and Rural Development Department, Government of West Bengal, India I was deputed by the director to provide technical assistance for piloting BPL census for 11th five year plan. The following assistances have been provided. 1. Training has been given to the field workers for collecting data in different

districts of West Bengal; 2. Supervision of the field work in selected villages of the districts of North

Bengal was made; 3. A manual entitled “The BPL census: Questionnaire Design, Survey

Methodology, Scoring Method, Cut off score – visited” was prepared and given to the paychayat and rural development department. This provided many suggestions on improvement of the schedule develop by Ministry of Rural Development for the Socio-Economic Survey 2010 and remedies of many problems to face in the field work.

4. For cost valuation of the collected data, a random sample from the respondents were drawn and given to them for revisits.

5. The following suggestions for analysis of data were made; An outline of the possible analysis of the data: Poverty alleviation programmes can be described as belonging to one of the programmes (i) programmes whose beneficiaries are self-selected (for example, MG National Rural Employment Guarantee Programme) or (ii) programme that are meant exclusively for predetermined target groups (for example, the Public Distribution System (PDS) for the provision of fair priced food and the Indira Awas Yojana etc.,). The criterion of targeting is,

65

most often, whether or not a household is below poverty line (BPL). Identifying such BPL households is crucial to implementation of targeted antipoverty schemes. Hence, BPL data and its quality are very much important and one should find ways and means to obtain relevant, reliable and timely information, necessary for identifying the BPL households.

Exclusion and Inclusion Errors: The important type of errors in BPL census one faces, involve errors that exclude poor household from the category of the poor and inclusion errors that include non-poor households in the category of poor. Some of the reasons behind such errors were identified as belonging to one or more of the following categories.

(i) Errors involved in the survey:- (a) In the questionnaire and in the investigation process; for examples,

respondents belonging to the joint family tend to report joint families as separate nuclear families in order to qualify separately for benefits. Thus, if benefits are to accrue, they will accrue not just to one household, but to different segments of undivided family, if the households were to be defined BPL. Such a situation may lead to inclusion of non-poor family in the poor category, if one can find that household size and total scores are correlated.

(b) Though the poorest people lack assets and are unable to borrow because of their poverty, the “type of indebtedness” parameter gives the highest score to a household that is not indebted. Such a household receives a higher score than the score assigned to a rich land lord who borrows only from commercial banks. Thus, a poor family has a chance of being excluded from the “poor category”. Distribution by types of funding agency may help in this direction for all those, but not for the poor family who does not get loans. This is possibly because of the existing system in the aggregation of 13-parameters to establish the absolute and relative position of each household with respect to poverty status in a village: An alternative method of developing aggregate score may be tried out as an exercise.

(c) Selection of the indicators and the scoring scheme for each parameter are also

crucial in identification of the BPL household. One should examine and identify the variables who are likely to be (i) misspecified or (ii) under specified or vague by and to the respondents. Replacement of the variables should be made, if possible, through examination of the available data. One may also find the contingent nature of the parameters like “status of children”, reasons for migration” which are not applicable to all households. One should try to for reconciliation of the kind of contingent happenings though some exercise.

66

(d) Another serious cause for exclusion and/or inclusion errors is the way cut off scores are set for each state, district and village. The state level cut off may be set at the level of the official planning commission poverty line. The determination of cut off scores for administrative divisions within the state (district, block, village for exampled is left to the state government. The aggregate cut off scores for the determination of BPL households could vary across the administrative entities. Therefore, it is necessary to make exercises on the determination of region-specific cut off scores.

Exercise of comparison of BPL census 2002 and the current BPL census 2007. A) With respect scoring parameters:

A comparison of data on scoring variables for each household in a village for the BPL censuses is difficult. It may be therefore advisable to compare aggregates at the village level such as number of households and its distribution by indicators to identify serious in consistencies in the two data sets, if any. To mention a few.

(i) Number of households and its distribution by size group of operational holding land;

(ii) Number of households and its distribution by type of dwelling unit; (iii) Number of household and its distribution by ownership of consumer durables; (iv) Number of households and its distribution by literacy rate; (v) Number of households and its distribution by child status; (vi) Number of households and its distribution by type of indebtedness; (vii) Number of households and its distribution by caste; (viii) Number of households and its distribution by average monthly income etc.,

B) With respect to Non-scoring parameter: -

(i) Examination of relationship between household size and the total score, (for each BPL, 2002, 2007);

(ii) Distribution of households by Social Group (separately for each BPL census, 2002 and 2007);

(iii)Distribution of households by land tenure and by size of operational land holdings; (separately for each BPL, 2002 and BPL 2007);

(iv) Distribution of households by average monthly income (separately for each BPL census);

Summary:-

(i) An exercise to examine the relationship between household size and total score; (ii) An exercise for developing alternative method of aggregation; (iii)An exercise on the identification of “misspecified”, underspecified” and vague

variables; (iv) To estimate proportion of households having the contingent parameters like

“reason of migration” etc.,

67

(v) Determination of the cut off scores at different hierarchical units and methodology;

(vi) Exercises of integration as mentioned in A and B above. 7. Participated at the 32nd Meeting of the National Statistical Commission (April 24, 2010):

At the invitation of National Statistical Commission at an interaction session with the commission at Kolkata. The following suggestions were offered.

(i) The system, though capable of capturing the wide variety of data generated on a given horizon and up to a given vertical distance of administration, it can not reach at the grass root level, as we shall see in the case of data requirement for local level planning.

(ii) Because of growing diversified requirement in view of the expanding economy leaning towards liberalization, and because of the change in the state-private sector mix, the country’s information need characteristic’ is changing and at the same time, because of the system’s dependence on the traditional records-mainly by a product of the administration of the age old administrative set up, data gap between the perceived need and the availability of it is gradually widening. Thus, the system is now partially unable now to produce relevant, comprehensive, accurate statistical information in some sectors.

(iii) The civil registration system should be extended for planning health and

family welfare programme at the local level as required by the 73rd and 74th amendment of the constitution.

(iv) Data collection and compilation work on Environment statistics is in a very

nascent state, and as such, efficient system for the collection, collection and compilation of data on environment statistics and development of environmental indicators are heart felt needs.

(v) Health sector requires, with respect to each disease, identification of major

chronic diseases, not only different types of cancer, heart-disease, hypertension, diabetes and Aids, but a strong desire of a statistical study is needed. The study would involve (i) data base for mortality, (ii) data base for prevalence and incidence; (iii) epidemiological and intervention studies; (iv) data base and literature on clinical disease; (v) statistical modeling at a population or molecular level..

(vi) The private sector is now replacing the public sector as a dominant force in the

economy and the statistical system must be redesigned to better meet its information needs.

68

(vii) Service Sector Activities are increasing and hence data must be collected on this tertiary sector of the economy.

(viii) There is a need for development of a computerized Data Base for service

sector along with appropriate methodology of data collection.

(ix) Legal/Legislative obstacles for collection, and collection of core statistics needs to be discussed.

(x) The system producing health statistics is totally decentralized and still

relatively week by Indian standards on incidence or prevalence of major diseases at the national level. It needs major overhaul and improvement.

(xi) The system being partly centralized, and partly decentralized, dependent on the

state government, state statistical agencies and other line departments, there must be provision of strong coordination between the statistical authorities at the state and at the centre. Regular meetings of the COCSSO must be held.

(xii) It may be seen that under the system, there exist multiple agencies for

collecting data on the same subject. For example, in case of employment, unemployment and under employment data can be made available through seven sources. In case of price statistics also, one may have different sources both at the state and at the centre level. If a data user faces the multitude of sources, he quite often finds that – (i) figures on the same subject differ by source; and (ii) often concepts, definitions, coverage and classifications made by different agencies for the same subject do not match. Therefore in such situations, one care must be taken to find some ways and means for reconcillation of the figures produced by different agencies. There should evolve some mechanism to control such differences between figures produced by different agencies.

(xiii) Recognizing the importance of environmental statistics as an emerging area

and it being multidisciplinary in nature, needs of standardization of relevant terminologies and concepts are felt so that people from different disciplines working together both at the state and at the center may be familiar with the uniform concepts and definitions. Otherwise, estimating number of species, for example, in biodiversity, may land up to an unreliable value. Here we will face new and peculiar difficulties such as the same plant having different names in different parts of the country. These have to be sorted out before biodiversity indices can computed. The same situation will apply to other sectors of the economic activity. A standardization programe should be under taken as an action programme.

69

(xiv) In surveys, non-sampling errors (e.g., under or over coverage, non-response) are more important than sampling errors. Non-response problems have worsened over the years. There is some theoretical research on rectification of these (e.g. Judith Ressler).

(xv) With regard to coverage problem, it has been observed that in case of ASI

frames, the coverage problem both in terms of under and over coverage exists, list of the population units being maintained by chief inspector of factories. With a recent learning towards a system of economic liberalization and with free flow of market research through the process of economic reforms, the industrial firms have lost some of the incentives to report their production to the Government of India. Some activity should be directed towards updating the ASI frame. The same problem of under coverage and over coverage exists in other surveys also. This problem should be addressed immediately.

(xvi) In a multistage design like that of NSSO, because of the quick changes in the

rural nature, one unit may turn out to be a non-respondent. Similar problems of non-response occur in many surveys. Existing methods of some imputation techniques may be put to use in handling non-sampling errors. Measurement errors also occur in many surveys; Existing techniques may be put to use in handling such errors. Along with the standard error of the estimate, estimation of non-sampling variance is also necessary.

(xvii) As the village and urban block level data on the number of enterprises and

workers as per the Economic Census (EC) are used as the sampling frame for selection of villages and urban blocks in the follow up enterprise surveys, necessary measures must be taken in the Economic Census (EC) to enhance the quality of data. A few years back, at the request from National Advisory Board of Statistics (NABS), a sampling methodology alternative to Economic Census was developed by a team of ISI scientists, namely Professor J.K. Ghosh, Professor Shibdas Bandyopadhyay, and Dr. Pulakesh Maiti for estimating certain parameters related to unregistered manufactures with high mortality rate. This method may be revisited and put to use in practice.

(xviii) In fact during April 1998 – July 1999, just before Rangarajan Commission

report came out, a research project entitled ‘strengthening local government in MP” was undertaken jointly with Price Water House India at the ISI. The project identified the items of information for local level planning, developed method of data collection, and developed a computerized data base. However a significant new follow up study has been made by Dr. Pulakesh Maiti.

70

The follow-ups were as follows: (a) Creation of a computerized data base for decentralized planning in the

district of Howrah, i.e., collection and use of statistics at gross roots levels for planning and decision making at the levels of panchayats;

(b) Development of a useful software for a computational data base the usefulness of the software was demonstrated to users;

(c) Preparation of fourteen different block wise Human Development reports

and one report for rural sector of the district;

(d) Preparation of one data based Gram Pnchayat Plan under the principal of decentralized planning.

(xix) While working under the above two projects, some additional basic statistics

other than those identified through 11th and 12th schedules of the 73rd and 74th amendments of the constitution at local level planning were needed to be created. They are as follows: (i) The nine fold classification of land use should be slightly enlarged to

cover two or three more categories such as social forestry, marshy and water logged land, from the point of decentralized planning;

(ii) The yield rate estimates based on scientifically designed crop cutting experiments under the general crop estimation survey (GCES) are not adequate to provide estimates below the district level; For decentralized planning and also with the introduction of National Agricultural Insurance Scheme (NAIS), a need is felt for assessment of yields at the level of block/teshil even at the panchayat level.

8. Working as Member of Expert Group in preparation of Annual and Perspective Plan of the District of Howrah under 11th 5 year Plan (April 2007-March 2012) In reply to the letter from the District Magistrate (Memo No.1283, dated 20.10.2006), I was nominated by the Director as a Member of the Expert Group in Preparation of Annual and Perspective Plan of the Howrah District, under 11th 5 Year Plan (No.D.O/12216, 26.11.2006) A Review of the work as a Member of the Expert Group can be summarized as follows. Planning for development invites 4 different types of activities – Formulation, Implementation, Monitoring during implementation and Evaluation on completion. To carry each of these activities, relevant, reliable and timely information is needed at every stage. The central idea of decentralization of planning is to make an effort to bridge the gap between the availability of resources at the grass root level and the subsequent planning process is built up on the basis of these information. Decentralization of planning needs

71

to be assessed from the perspective of functional, administrative and financial devolution and the extent of people’s participation. Additional responsibility has been assigned to the local Governments by the 73rd and 74th Amendments to the Constitution of India. Writing a Monograph: The monograph entitled Statistical Information System for Decentralised Planning in the district of Howrah written by myself has discussed such issues for development of database for decentralized planning as

a) identification of the items of information needed for decentralized planning; b) the level at which data may be required; c) designing of the formats of data collection, collation and compilation; d) methodology of data collection; e) identification of data collecting agency and f) specification of output formats amenable to computerised database.

It has also discussed a review of the existing information gathering systems including the Statistical Information System developed at the Indian Statistical Institute, Kolkata under the ISI-PWI project [Maiti et al (1999)]. Volume of Information: Volume of information needed for creating the database, different types of work, the necessary man power, time and cost has also been indicated in the same monograph. A few recommendations on

a) the present data recording system, b) need for availability of electronic media at the level of different local bodies, c) need of training programme to be imparted at the local level resource persons

for updating data were suggested there. A description of the Statistical Information System to be developed for the district of Howrah was also outlined in the monograph. Description of Error Analysis: Assessment in the quality of data is needed to avoid creating an undue expression that data so unreliable as to be of no use. It is necessary to provide some idea of the reliability of results. Care was taken to setup controls through proper :

a) schedule design; b) survey design; c) training to the survey management group and d) to data processing and other personnel involved in the operation.

It became still necessary to examine that controls were effective and results with desired accuracy have been achieved. Identification of the errors were made through

a) internal evidence and external administrative check; b) internal evidence and subjective expert views; and c) internal evidence alone.

The errors detected and corrected appeared to be belonging to one of the following categories: a) unacceptable because the information appears to be impracticable; b) unacceptable because of logical inconsistency;

72

c) unacceptable because of conceptual mistakes; d) unacceptable because of they remaining blank; and e) unacceptable because of misspecification of identity of the basic units.

Copying errors while entering raw data to MS Excel file and further loading the MS Excel file data into the RDBMS were detected and corrected

Creation of the database: As startup activities for creating the RDBMS the following studies were made: 1. System Requirement Study (SRS); 2. Context Analysis and Design (CAD); 3. Data Flow Diagram (DFD).

RDBMS is a computerised linkage in third normal form and consists of the tables which are broadly classified into two categories namely, item-wise tables and hierarchy-wise tables. The criteria considered when designing tables in the RDBMS included:

a) information rule; b) guarantee of access rule; c) systematic treatment of null values; d) online catalogue; e) view updating rule; f) high level insert, update; g) physical and logical data independence; h) integrity independence; i) distribution independence; j) non-subversion rule; k) data non-redundancy; l) consistency in data stored

Uploading the data: Before uploading the data in to the database, the following necessary work were done:

a) manual correction of Area Master names in the accumulated MS Excel files; b) rectification of cell formats in the MS Excel file from text to number and vice

versa, percentage to number, decimal to number, accounting to numberetc.; c) correction of area names as per Area Code List (ACL); d) the cell formats were corrected for extraction of data without error from MS Excel

file to the tables of RDBMS; e) the uniqueness in names was guaranteed to maintain relational integrity among

the tables of the RDBMS; f) MS Excel data files contained some alphanumeric values, which were converted

to numeric values before loading into the database, as the basic criteria of uploading the data is that data should be numeric in nature only;

However, some corrections were not possible to be made as data in dates, fractions and scientific formats are in irreversible form. Data were stored in the tables of the RDBMS and for that necessary instructions were developed to append the data into the tables of RDBMS. Preparation of 157 booklets of Resource Mapping for 157 Gram Panchayats of the district(Data Content of the Statistical Information System):

73

For processing the data and compilation of the above booklets required development of an application software. Fundamental work in preparation of the proposed application software included :

a) designing of forms by adding objects like grid view, frames, text boxes, list boxes, combo boxes, commands buttons on them;

b) layout/design of data entry forms in consistent with Area Master hierarchy was made.

Separate data entry forms were developed for different items of information to be filled-in from the manual survey schedule. Development of the application software model needed separate entry forms for 12 blocks of the survey schedule. With the help of the developed application software, Gram Sansad wise 34 basic reports were prepared. They were as follows: i) Educational Amenities I ii) Educational Amenities II iii) Medical Amenities iv) P & T and Internet Amenities v) Communication Amenities vi) Financial Institution Amenities vii) Market/Haat Amenities viii) Office Amenities ix) Service Centre Amenities x) Power Connection Amenities xi) Power Disconnection Amenities xii) Power Non-connection Amenities xiii) Drinking Water Amenities I xiv) Drinking Water Amenities II xv) Irrigation Water Amenities I xvi) Irrigation Water Amenities II xvii) Both Irrigation & Drinking Water Amenities I

xviii) Both Irrigation & Drinking Water Amenities II

xix) Stock of Animal Husbandry xx) Educational Institutions Faclity xxi) Development Project Beneficiaries xxii) Land and Water Body Use – I xxiii) Land and Water Body Use – II xxiv) Cultivated Land – Main Crops xxv) Cultivated Land – Pulses xxvi) Cultivated Land – Oil seeds xxvii) Cultivated Land – Other Produce xxviii) Cultivated Land – Spices xxix) Cultivated Land – Vegetables xxx) Cultivated Land – Cash crops xxxi) Health Service Available xxxii) Health Scheme Beneficiaries I xxxiii) Health Scheme Beneficiaries II xxxiv) Non-conventional Energy Source

Preparation of Human Development Reports: The 73rd and 74th Amendments to the Constitution in India has given rise to acceptance of decentralized planning process and speaks of the bottom-up approach to planning, where the felt needs of the people from the grass root level are assessed. For this, Inter-Block and Intra-Block variation need to be brought out. The work for preparing Human Development Reports started with the creation of above database and completion of compilation of 157 above booklets of resource mapping for 157 Gram Panchayats. Each of the 14 Blocks of the district was separately examined from the point of its attainment in the level of human development based on the above primary data collected and processed at the different levels of hierarchy including Gram Sansad, the lowest level.

74

In the basic reports the extent of land use and cropping pattern, the stock of animal husbandry, sources of drinking water and water for irrigation, rural electrification etc. have been displayed for each of the 14 blocks. These have made possible to prepare 14 Block-wise Human Development Report separately entitled Block-wise Human Development Report in the district of Howrah, West Bengal. Also has come-up the single Human Development Report for the rural sector of the district as whole entitled Inter-Block variation with respect to Human Development in the Rural Sector of the district of Howrah. Reports Published Under the Above Activities as a Member of Expert Group in Preparation of Annual and Perspective Plan of the District of Howrah Under 11 Five Year Plan: The following are the total number of publications under the above project: 1. Statistical Information System for local level planning by local bodies in the District of Howrah. A monograph published by Howrah Zilla Parishad; 2. 157 booklets of Resource Mapping for each of 157 Gram Panchayats of the District of Howrah. published by Howrah Zilla Parishad; 3. Inter-Block variation with respect to Human Development in the Rural Sector of the district of Howrah. The theme conceived and supported by Howrah Zilla Parishad; 4. 14 Human Development Reports at the Block level for each of the 14 Blocks entitled Block-wise Human Development Report in the district of Howrah, West Bengal. The theme conceived and supported by Howrah Zilla Parishad.

Design and Development of Computer Software as a Member of the Expert Group in Preparation of Annual and Perspective Plan of the District of Howrah under 11th Five Year Plan: Startup Activities:

1) System Requirement Study (SRS) was made to gather an in-depth knowledge about the proposed system i.e Statistical Information System for Rural Decentralised Planning (SISRDP).

2) Context Analysis Design (CAD) was prepared to depict the information inflow and out-flow from the proposed system (SISRDP).

3) Preparation of Data Flow Diagram (DFD) to figure out the processes required to generate the outputs (reports etc) from the given inputs (Area Master Names and Codes, detailed survey data etc.).

4) Demarcation of Automation Boundary in the developed DFD. 5) Development of Area Master form ( to enter Area Master names and their codes) and

detail data entry screens for filling detailed data from the manual schedule (survey form).

Fundamental Work in preparation of proposed Application Software:

1) Forms were designed by adding objects like grid view, frames, text boxes, combo boxes, list boxes, check boxes, command buttons on them. The properties of these objects are then bound to each and every data entry form (both in Add and Update

75

mode), report generation screen. Logical codes were written to populate objects like combo boxes for effective date, list boxes for various Area Master names like Block (Administrative Block), Gram Panchayat, Gram Sansad etc. Grid view to show data fetched from tables of the database. The Area Master names are filtered basing on names selected in upper hierarchy. For example, when a Block is selected from the drop down box (combo box), only Gram Panchayats of that Block get populated.

2) Layout/design of data entry forms are developed in consistent with Area Master hierarchy (like Gram Sansads under Gram Panchayats which is under Block/Panchayat Samiti). Separate data entry forms were developed for different items of information to be filled-in from the manual survey schedule.

Important Feature of the Application Software:

1.The application software has been so developed that in future new survey data can be accommodated, edited/updated and the reports (basic, summarized as well as query based) can be generated without any further modification in the database (RDBMS) and the application software (SISRDP). This would help one append, revise, correct and generate reports according to their requirement. 2.SISRDP can be made dynamic in the sense that it can be applied for another hierarchical stage, say for another District, with some modifications, as an additional hierarchy for the district has to be provided. 3The programs developed needs revision to reach the additional hierarchy i.e. the District. All programmes such as

• Area Master form/screen; • detailed survey data entry forms in addition/updation mode; • reports (basic reports and summarised) as well as other reports at the

Block and District level need to be modified so that District can be counted as a member in the area hierarchy. 4.This additional work to make the application software workable (for any district) should not take more than six months.

Development of the application software using VB6, VB6 SP6, MS Projects includes separate Area Identification Masters for the following:

i) Sub-division ii) Block/Panchayat Samity iii) Thana iv) Gram Panchayat v) Mouza vi) Village vii) Gram Sansad

and separate Entry Forms for 12 blocks of the Survey Form.

76

They are as follows: i) Area Identification through selection of Sub-division name, Block name, Thana, Gram

Panchayat name, Mouza name, Village name and Gram Sansad from their respective drop down box. This also includes no. of families, males, females and children residing in the Gram Sansad.

ii) Infrastructure Available – here also Area IDs need to be selected from available drop down list boxes, then distance from the respective Gram Sansad to the Available Infrastructure in KM are to be mentioned.

iii) Land & Swamp Use – Area IDs are to be selected first. Then area of land and swamp under different uses within the Mouza can be mentioned in hectors.

iv) Domestic Animal/Bird - Area IDs are to be selected from respective drop down boxes. Then number of domestic animals/birds in the category of male, female and calf/kid in the Gram Sansad can be entered.

v) Irrigation & Drinking Water Source - Area IDs are to be selected from respective drop down boxes. Then number of irrigation and drinking water source in the category of only irrigation, only drinking water and source under both use within the Gram Sansad are to be mentioned.

vi) Land under different Crops produce - Area IDs are to be selected from respective drop down box. Then area of land under Principal Crops, Pulses, Oil seeds, Other Produce, Vegetables, Masala and Cash Crops categorized into Kharif, Rabi and Summer crop land within the Mouza are to be entered.

vii) Educational Institutions - Area IDs are to be selected from respective drop down boxes. Then educational institutions available within the Gram Sansad can be entered in such categories as, name of school, type of school as per level, type of school as per structure, number of boys and girls students and number of male and female teachers in it.

viii) Health Indicators – here also Area IDs are to be selected from available list in drop down boxes. Then data on Health Sub-centre, Homeopathy Centre and Ayurvedic Centre each categorized into name and number of beds in each centre within the Gram Panchayat can be entered.

ix) Health Scheme Beneficiaries – select Area IDs from available list in drop down boxes. Then number of beneficiaries under Family Planning, T.B Control, Leprosy Prevention, Blindness Prevention, Malaria Prevention and different sub categories of Mother and Child Welfare Project within the Block can be entered.

x) Development Project Beneficiaries - here also Area IDs are to be selected from available list in drop down boxes. Then number of beneficiaries in different development projects under such categories as name of Project, Amount Allotted for the project, Actual Expenditure made, Project Start Date, SC, ST, OBC, General and total beneficiaries within the Gram Sansad can be entered. Each type of beneficiaries are further sub categorised into Male, Female and total beneficiaries.

xi) Renewable Energy Source – first Area IDs are to be selected from available list in drop down boxes. Then number of Bio-gas, Solar Energy and Wind Mill within the Block are to be entered.

xii) Electric Connections – first the required Area IDs are to be selected from respective drop down boxes. Then number of electrified, previously electrified presently not, non-electrified and total of them under categories Domestic, Commercial, Cottage Industries,

77

Very Small Industries, Medium Industries and Irrigation in the Gram Sansad are to be entered.

During system requirement study it was observed that if the distance from Gram Sansad to nearest Primary School is 0 (in Infrastructure Available), then there must be one entry for primary school in Educational Institutions within Gram Sansad. However,such cross validations were not carried out since such inconsistent data were modified later through Edit option. The following programs were designed and developed in the Application Software: Design and development of Sub-division Area Master form with necessary template for data addition/updation and retrieval.; Design and development of Sub-division Area Master form with necessary class (template) for data addition and retrieval.; Design and development of Block/Panchayat Samiti Area Master form with necessary class for data addition and retrieval; Design and development of Thana Area Master form with necessary class for data addition and retrieval.; Design and development of Gram Panchayat Area Master form with necessary class for data addition and retrieval;. Design and development of Mouza Area Master form with necessary class for data addition and retrieval.; Design and development of Village Area Master form with necessary class for data addition and retrieval. Design and development of Gram Sansad Area Master form with necessary class for data addition and retrieval; Design and development of Integrated Area Master form with automatic data populating, addition and updation facilities. The previously developed class files were used with some changes made to them to incorporate thana name in Block master class file and village name in Gram Sansad Master class file. Changes in Area Master hierarchy from Sub-division – Block – Thana – Gram Panchayat – Mouza – Village – Gram Sansad to Sub-division – Block – Gram Panchayat – Gram Sansad were made making thana optional and Village independent in database tables and Area Master form. Design and development of Detailed Surveyed information entry form with 12 tabs for 12 blocks of the manual survey form:. The Area Identity exists in the tab named “Area Identification”; for such other tabs as respective Gram Sansad, Mouza, Gram Panchayat, Block name gotautomatically displayed. Like when entry will be made in “Infrastructure Available” tab, name of the respective Gram Sansad will automatically be shown in the top left corner heading;Similarly, when an entry will be made in the “Land and Swamp Use” tab, name of the respective Mouza will automatically get displayed in the top left corner heading

78

Some Operational Features:

a) Area Master selection from drop down box (where changes have been made in area master hierarchy); b) no. of families in a Gram Sansad, no. of males, no. of females, no. of children and automatic calculation of population in a Gram Sansad.; c) Distance from respective Gram Sansad to 51 infrastructural facilities segregated into 7 distinct groups (Education, Financial Institutions, Communication, Health, Market/Haat, P&T and Internet and Office and Service Centre); d) Area under different Land & Swamp use Mouza wise in hectors; e) No. of different domestic animals & birds Gram Sansad wise; f) Gram Sansad wise no. of different source of irrigation and drinking water; g) Mouza wise land under Principal crops, Pulses, Oil Seeds, Other Produces, Vegetables, Spices, Cash crops in Khariff, Rabi and Summer seasons; h) Educational Institutions in Gram Sansad mentioning details of same of school, type of institutions, structure of institution as per tier and as per construction, number of Boys & Girls students and number of male and female teachers; i) Gram Panchayat wise Health Indicators mentioning Type of Centre (which will any one of Health Sub-centre, Homoeopathy Centre and Ayurvedic Centre), Name of Centre and No. of Beds in the center; j) Block wise Number of Health Beneficiaries which has been segmented into 3 categories viz a) Beneficiaries under Family Planning Program, b) Beneficiaries under different disease prevention projects like TB Control, Blindness Prevention, Malaria Prevention, Leprosy Prevention; and c) Mother and Child Welfare Program in which beneficiaries under DPT, TT, Polio, BCG, Chicken Pox, Anemia, Vitamin ‘A’ deficiency and Total Sanitation Project are mentioned. k) Gram Sansad wise Development Project Beneficiaries in which name of Development Project, allotted amount, actual expenditure made, project start date, beneficiaries under different caste like SC,ST,OBC general and total of beneficiaries under different caste are mentioned l) Block wise number of renewable energy sources like Bio-Gas, Solar Energy and Wind Mill are mentioned

m) Gram Sansad wise number of Electrical Connections, Disconnections and no-connections in domestic, commercial, cottage Industries, small scale industries, medium scale industries and in irrigation are mentioned.

Preparation of Statistical Distributions of a number of Statistical variables:

A number of Application Software Programmes have been developed to process and generate the statistical distributions of different variables. This has resulted in to develop ment of 105 software programmes.

79

24. A Plan of Future Research Work in the Next Following Years:

The survey process consists of the processes namely (a) sample selection,(b) response from a sample member,(c) measurement of those agreeing to participate and (d) imputation for those selected, but unwilling to participate, which are assumed to be stochastic, Another source of survey error occurs when the sampling frame is imperfect. Imperfection may be thought of as the presence of certain kinds of matching problem between the set of frame population and target population elements. The match up of frame and target populations is established by the indicator variable, which is viewed as deterministic rather than stochastic, since sampling is often conditioned on the existing sampling frame, thus making the linkage predetermined and fixed in practice.

Given enough evidence of different sources of error with their interrelationships, it is important to develop total error model for survey estimates. Attempts to reduce or control errors of one type may have adverse effects on some other components of total error. It is interesting to examine the nature of total error model, i.e., those that accommodate several sources of error in estimating more complex parameters than the total or mean only. Development of appropriate cost models would also be necessary in the context of total survey design. The study of non-sampling errors and its effect on different estimators – linear as well as non-linear would worth studying to the coming days.

80