assumption 2

22
EGERTON UNIVERSITY AFRICAN INTERNATIONAL COLLEGE CENTRE SCHOOL OF ARTS AND SOCIAL SCIENCES DEPARTMENT OF PEACE, SECURITY AND SOCIAL STUDIES COURSE TITLE: RESEARCH METHODS COURSE CODE: SOCI 702 ASSIGNMENTS PRESENTED BY: NAMES REGISTRATION NUMBER SISCO CHES CHEROP AM/AICO/00001/11 TO: LECTURER’S NAME: DR A. KISILU 1

Upload: onyango-okoth

Post on 24-Oct-2014

60 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Assumption 2

EGERTON UNIVERSITY

AFRICAN INTERNATIONAL COLLEGE CENTRE

SCHOOL OF ARTS AND SOCIAL SCIENCES

DEPARTMENT OF PEACE, SECURITY AND SOCIAL STUDIES

COURSE TITLE: RESEARCH METHODS

COURSE CODE: SOCI 702

ASSIGNMENTS

PRESENTED BY:

NAMES REGISTRATION NUMBER

SISCO CHES CHEROP AM/AICO/00001/11

TO:

LECTURER’S NAME: DR A. KISILU

PURPOSIVE SAMPLING

1

Page 2: Assumption 2

Sampling is an essential process in research especially to large populations or a population that is not easily accessible. Sampling can as well be important to a study whose result is meant for estimations and forecasting.

There are many techniques of sampling, which can be found within probability and non-probability samplings. The probability sampling there is a chance more than zero of an event being selected for the study, while in the non-probability this chance is zero. That means the difference between nonprobability and probability sampling is that nonprobability sampling does not involve random selection and probability sampling does.

Hence nonprobability samples cannot depend upon the rationale of probability theory. At least with a probabilistic sample, the odds or probability are known that the population has been represented well. The confidence intervals can be estimated for the statistic. With nonprobability samples, it may or may not represent the population well, and it will often be hard to know how well it has been done. In general, researchers prefer probabilistic or random sampling methods over nonprobabilistic ones, and consider them to be more accurate and rigorous.

However, in applied social research there may be circumstances where it is not feasible, practical or theoretically sensible to do random sampling. Here, we consider a wide range of nonprobabilistic alternatives. The nonprobability sampling methods can be subdivided into two broad types accidental or purposive. Purposive sampling technique allows a researcher to use cases that have the required information with respect to the objective of his/her study. Most sampling methods are purposive in nature because the sampling problem is usually approached with a specific plan in mind. The most important distinctions among these types of sampling methods are the ones between the different types of purposive sampling approaches.

In purposive sampling, a sample is made with a purpose in mind. Usually there would be one or more specific predefined groups being sought. For example, have you ever run into people in a mall or on the street who are carrying a clipboard and who are stopping various people and asking if they could interview them? Most likely they are conducting a purposive sample. They might be looking for huge (obese) females between 30-40 years old.

They size up the people passing by and anyone who looks to be in that category they stop to ask if they will participate. One of the first things they're likely to do is verify that the respondent does in fact meet the criteria for being in the sample. Purposive sampling is very useful for situations where a targeted sample needs to be reached quickly and where sampling for proportionality is not the primary concern. With a purposive sample, it is likely to get the opinions of the target population, but it is also likely to overweight subgroups in the population that are more readily accessible.

2

Page 3: Assumption 2

The following methods are considered subcategories of purposive sampling methods. There may be sampling for specific groups or types of people as in modal instance, expert, or quota sampling. We might sample for diversity as in heterogeneity sampling. Even capitalize on informal social networks to identify specific respondents who are hard to locate otherwise, as in snowball sampling.

Advantages This method has the following advantages:

(i) Evan a sample can be representative thus if prejudices and bias have been avoided, even a small sample can be representative.

(ii) In this method, if the samples have been selected cautiously, the purpose of research if fulfilled without botheration.

(iii) If the investigator has an idea about the universe, he shall note which units are useful for the research. In this method, the selector has supreme say and he selects all those units that are useful for the study. As a result of this, all relevant and useful units are given place.

Disadvantages i) This methods starts with the assumption that the investigator has complete

knowledge of the universe in advance but in normal course, it is impossible. Because of this lack of knowledge, it is not possible to make scientific selection.

ii) The restrictions and the controls that are imposed in the selection of the units are not very effective. That is why Neyman has said that “this method is hopeless”.

iii) In this method, because of bias and prejudices, inaccuracies creep in, in fact, purposive sampling is based on hypothesis in which it is not possible to test through practice, that is why mistakes that have crept in the selection are not easily detected.

INTERVIEWS

3

Page 4: Assumption 2

An Interview is an oral administration of a questionnaire or an interview schedule. Interviews are therefore face-to-face encounters. The researcher needs to obtain accurate information. Focus is on research interviews which are a kind of diagnostic interview in which the cause of a problem is sought and analysis in detail.

The mention of interviews creates a flutter in many stomachs. There are not many who can remain unaffected by the idea of facing an interview. Actually, we pass through interview situations very frequently without being aware of them. For example, when waiting in a railway hall for your scheduled train you could start a conversation with the man in the adjacent chair. You may exchange names, destinations of your travel, occupations and other such information. In   another case when you visit your problems in order to diagnose and treat. Several such situations happen all the while and we deal with them effortlessly, while facing a formal interview for administration or recruitment we are very aware of the situation. Even in research interviews need a lot of preparation and we can perform at them with role-playing practices.

Objectives of InterviewCollecting information through face-to-face contact: An interview is a direct technique of collecting data by the interviewer. Its advantage lies in the fact that a great deal of information needed in social and scientific research or assessment can be collected from the respondents only by direct questioning.

Formations of hypothesis:An interview is an exploratory device to help identify variables and relations to suggest a hypothesis and to guide other phases of the research this is applicable only to interviews for research

Collecting information about qualitative facts:Social facts are basically in nature found in the forms of ideas, feelings, views, faith, and beliefs. These social facts can be individual as well as collective. The interview method is the best way to collect qualitative facts.                                                               

Collecting Additional InformationBy this method, we can collect additional information from normal respondents than we often get through schedule or questionnaire. There are certain persons who suggest or are capable of providing additional information and this can be done most effectively through interview method. This is possible because of a direct dialogue with the respondents.

Importance of the interview method:This is the best method for getting information about feelings:No other method of social research provides better information on feelings, emotions, and sentiments than an interview.

4

Page 5: Assumption 2

Useful for securing information from persons at all levels:The questionnaire method is useful only for the educated. Sometimes, questionnaires cannot elicit the required response from persons of different levels of intelligence. This drawback is overcome by the interview method.

Method of inter-stimulation:As a result of two or more persons coming in contact with each other during an interview they stimulate each other mentally. This helps the respondent to come up with the answers that are required of him. By observing the respondent closely, an interviewer can ascertain some aspects of the respondent’s behaviour.

Verification of information:Information collected through questionnaires and observation cannot be verified easily. But in an interview, it is possible to verify the information that has been collected from sources like the bio data. Thus an interview is a vital method of data collection in research, a powerful device to obtain options, and a popular evaluative technique for whatever purpose the interview is being held.

Features of Interview It is close contact or an interaction including dialogue between two or more persons.There is a definite object of the interview such as knowing the views and ideas of others.There is face to face contact or primary relationship between the individuals.

Types of interview forms in researchInterview on the basis of number of respondents

Group interview Individual interview          Interview on the basis of form Structured interview Unstructured or Informal interview

Over the years different techniques of interviewing have been designed by companies government organizations, autonomous bodies, and private organizations.

MeritsIn the field of social research, particularly when the study deals with the personal life of the respondent, an interview is a popular method. The data required in such studies can be provided only by the individual being interviewed. The method has the following advantages:

Possible to study events not open to observationThere are events which are not known to anybody expect the respondent and which cannot be studied through observation. Through the interview method it is possible to study such events.

5

Page 6: Assumption 2

Possible to abstract factors like Attitude, Feelings, Emotions, and ReactionsAn interview helps to study abstract and intangible personal factors like feelings, emotions, faith and belief. Social phenomena influence all these intangible personal factors. An interview makes it possible to learn influence of the social phenomena on these factors. Riots, public meetings, festivals and so forth are such examples.

The information gathered is ReliableThe information is gathered through an interview is reliable if the technique adopted is pre-tested and revised to eliminate ambiguities and inappropriate wording

Useful for all population segmentsThe interview method is useful for all segments of the population, unlike a questionnaire which is useful only for the literate population.

De-meritsThe Data Collected is of Doubtful characterThe possibility of the interview being biased or prejudiced makes, the data collected in an interview suspect.

Too much Dependence on memoryOften, it is not possible for the interviewer to record things during an interview. He has to depend on his memory which might not serve him well later.

Specified knowledge is always not possibleThe interviewer has to be an expert in human psychology and must understand human behaviour. She/He should have a great personality and initiative. These qualities are not always available with every interviewer and the interpretation of information suffers several drawbacks.

Play of complexesIf the interviewee feels discomfited by certain questions, it colors his answers. And if he/she is overlooked, his enthusiasm dampens.

Difference in the social background of the interviewer and the intervieweeVery often the interviewer and the interviewee are not from the same social back-ground, this creates a difference in their respective norms and values. As a result they may not agree on some points. This difference violates the objective of collection of data.

6

Page 7: Assumption 2

ACTION RESEARCHWhat is action research?Discussions about the history of action research usually begin with the work of Kurt Lewin in the 1940s. Whilst most literature concerns educational action research (Dick 2006) an action research approach is also sometimes employed in community and organization research. There are many definitions- this one is from McNiff (2002).

“Action research is a term which refers to a practical way of looking at your own work to check that it is as you would like it to be. Because action research is done by you, the practitioner, it is often referred to as practitioner based research; and because it involves you thinking about and reflecting on your work, it can also be called a form of self-reflective practice.

The idea of self reflection is central. In traditional forms of research – empirical research – researchers do research on other people. In action research, researchers do research on themselves. Empirical researchers enquire into other people’s lives. Action researchers enquire into their own” (McNiff 2002).

Understanding Action ResearchAction research is a process of deep inquiry into one's practices in service of moving towards an envisioned future, aligned with values. Action Research is the systematic, reflective study of one's actions, and the effects of these actions. The researchers examine their work and seek opportunities for improvement.

As designers and stakeholders, they work with colleagues to propose new courses of action that help their community improve work practices. As researchers, they seek evidence from multiple sources to help them analyze reactions to the action taken. They recognize their own view as subjective, and seek to develop their understanding of the events from multiple perspectives. The researcher uses data collected to characterize the forces in ways that can be shared with other researchers /practitioners. This leads to a reflective phase in which the designer formulates new plans for action during the next cycle.

7

Page 8: Assumption 2

Action Research is a way of learning from and through one's practice by working through a series of reflective stages that facilitate the development of a form of "adaptive" expertise. Over time, action researchers develop a deep understanding of the ways in which a variety of social and environmental forces interact to create complex patterns. Since these forces are dynamic, action research is a process of living one's theory into practice.

The subject of action research is the actions taken, the resulting change, and the theory of change that is held by the persons enacting the change. While the design of action research may originate with an individual, social actions taken without the collaborative participation are often less effective. Over time, the action researcher often extends the arena of change to a continually widening group of stakeholders. The goal is a deeper understanding of the factors of change which result in positive personal and professional change.

This form of research then is an iterative, cyclical process of reflecting on practice, taking an action, reflecting, and taking further action. Therefore, the research takes shape while it is being performed. Greater understanding from each cycle points the way to improved actions.

8

Page 9: Assumption 2

Goals of Action Research include: A deep understanding of practice and the development of a well specified theory

of action; An improvement in the community in which your practice is embedded through

participatory research. The improvement of professional practice through continual learning and

progressive problem solving;

Action research as a method is scientific in which the effects of an action are observed through a systematic process of examining the evidence. The results of this type of research are practical, relevant, and can inform theory. Action Research is different than other forms of research as there is less concern for universality of findings, and more value is placed on the relevance of the findings to the researcher and the local collaborators. Critical reflection is at the heart of Action Research and when this reflection is based on careful examination of evidence from multiple perspectives; it can provide an effective strategy for improving a community’s livelihood, an organization's ways of working and the whole operations climate among others. It is the process through which people learn.

Developing Action Research Questions: A Guide to Progressive InquiryThe questions asked by action researchers guide their process. A good question will inspire one to look closely and collect evidence that will help find possible answers. What are good examples of action research questions? What are questions that are less likely to promote the process of deep sustained inquiry? The best question is the one that will inspire the researcher to look at their practice deeply and to engage in cycles of continuous learning from their everyday environment. These questions come from a desire to have practice align with values and beliefs. Exploring these questions helps the researcher to be progressively more effective in attaining their goals and developing professional expertise.

Good questions often arise from visions of improved practice and emerging theories about the change that will move the researcher closer to the ideal state of society. When stated in an if/then format, they can take the shape of a research hypothesis. If I [insert the action to be taken], how will it affect [describe one or more possible consequences of the action]?

Recognizing Weak Action Research QuestionsQuestions with known answers where the goal is to "prove" it to othersFor example, suppose a person has been holding family math night for years and sees an effect on parent participation. A weak question for action research would be: Will holding a family math night increase parent participation? This might be a useful evaluative research question where a controlled study could be set up to explore the connection. But evaluative research is different than action research. Action research is an experiment in design, and involves implementing an action to study its consequences

9

Page 10: Assumption 2

Questions that can be answered yes or noGenerally these are questions that will not encourage paying attention to the many nuances of the setting and the social interactions. Although, like any guide, while some yes/no questions can provide direction, it is often helpful to think about ways to transform the question into a different format. For example: Will the introduction of project-based learning lead to more student engagement? This .might be reworked to How will the introduction of project-based learning affect student engagement in my classroom? The first one, the researcher can answer the question with yes (an outcome that they might have expected). The second question guides them to look for the possible mechanism of project-based learning (maybe ownership, collaboration, or self-assessment) that have been found to be related to increased engagement. 

Questions that can be answered by reading the literatureWhat does community of practice mean? This might be a question that the researcher needs to answer, and can do so by reading more readily than by engaging in action research. A better formulation for action research might be: How will increasing the time for teacher collaboration in grade level teams affect the development of a community of practice at our school?"

Sharing your Action Research with Others:One of the strongest acts of leadership can be the act of writing of sharing knowledge and insights gained. Writing enables contribution to the body of knowledge that exists beyond the researcher. The final report serves the purpose of sharing the knowledge gained through action research with others in a community of practice. Action Researchers will need to decide what to write and to whom to write.

10

Page 11: Assumption 2

ASSUMPTIONS AND LIMITATIONSAssumptionsAssumptions are any important fact presumed to be true but not actually verified (Gay, 1976). They do not need testing, unlike the hypothesis. Some researchers spell out their assumptions but others do not. They form statements by the researcher that certain elements of the research are understood to be true. While assumed, they should still be explicitly stated in the body of the research.There are many assumptions that operate in every research study; these assumptions reflect the researcher's beliefs about the theory, the methodology, and the substantive phenomenon under investigation (Brinberg and McGrath, 1985; Fain, 2004). Although these core assumptions are most easily translated into a quantitative framework, they are also suitable (with a few modifications) for naturalistic modes of inquiry.

When researchers pause to reflect on the theoretical, methodological, and substantive aspects of their research, they knowingly or unknowingly, employ two criteria: veracity and sensitivity. Veracity refers to adherence or conformity to facts or truths; it is a judgment about the appropriateness or accuracy of a theory, method, or substantive phenomenon. Sensitivity refers to perception or susceptibility to prevailing attitudes, feelings, or circumstances; it is an evaluation of the receptivity or responsiveness of a theory or method. Because a substantive phenomenon is what it is, the sensitivity criterion is not relevant for judging the phenomenon under investigation. The most common types of research assumptions include: Assumptions are made about the theory under investigation, the phenomenon under investigation, the instrument, the methodology, the analysis, the power to find significance, the participants in the study, and the results. These are further discussed below:

The Instrument AssumptionThe researcher assumes that the variables under investigation are measurable (sounds easy: gender, age, and others, but how can one measure happiness?), and the instrument being used is a valid and reliable instrument to measure those variables. These statements can be made: that the interviewee answered the interview questions accurately, that the respondents rated some situation truthfully, that the attitudes of a group of respondents are stable, and that the subjects of the study all belong to the low socio-economic status group.

Phenomenal AssumptionsBefore beginning the study, the phenomenon under investigation must be clearly defined, and it must be measurable. It is assumed the variables have been clearly defined and are measurable.

The Methodology AssumptionThe researcher assumes the methodology is appropriate to the problem being addressed and the purpose of the study. For example, quantitative analysis is rarely appropriate to address how or why questions.

11

Page 12: Assumption 2

The Analysis AssumptionEvery statistical procedure has certain requirements. For example, most parametric analyses (e.g., Pearson correlations, ANOVAS) require normally distributed data. Therefore, you, as the researcher, assume that the data will be normally distributed. If the data are NOT normally distributed, then you might consider using a nonparametric procedure such as Spearman Rank instead of the Pearson correlation coefficient.

The Power to Detect AssumptionBefore conducting the analysis, the researcher assumes that the analysis selected and the sizes of the sample are sufficient to detect significant differences/relationships if they exist in the population.

The Participants AssumptionIn order for a study to be valid, the participants must be representative of the population, must be willing to participate in the study, and must respond to questions honestly or participate without biasing the study results (that is not behaving differently than they would were they not participating in a research study).

The Results AssumptionOnce your analysis is complete, you assume that the results are generalizable beyond the sample being studied. Finally, it is assumed that the results of the study will be relevant to stakeholders. This is the most compelling assumption: that the results will be meaningful.

LimitationsLimitations are those elements over which the researcher has no control. In most instances, any assumption you make becomes a limitation, while delimitations are those elements the researcher can control. They form phrases or aspects of the investigation which may affect the result of the study adversely but over which you have no control. It must be stated honestly.  Every study, no matter how well it is conducted, has some limitations. Therefore it is not reasonable to use the words "prove" and "disprove" with respect to research findings.  It is always possible that future research may cast doubt on the validity of any hypothesis or the conclusions from a study.  

Limitations of Correlational StudiesCorrelational research also has the same limitations as case studies.  Correlational research merely demonstrates that we can predict a variable from another variable.  It is demonstrating that two variables are associated.  However, two variables can be associated without there being a causal relationship between the variables.  We cannot make causal conclusions from correlational findings because we cannot rule out all alternative explanations for correlational findings. 

12

Page 13: Assumption 2

Thus, making causal conclusions from correlational findings is a logical error.  If we find that A is associated with B, it could mean that A caused B, B caused A, or some third variable caused both A and B without there being any causal relationship between A and B.  Even if we could rule out one of the possible relationships (e.g., B caused A), we cannot rule out all alternative explanations from correlational studies. For every correlational study, there is the possibility that some third variable caused the two variables without there being a causal relationship between the variables.

Correlational research may also have limitations with respect to the generality of the findings.  Perhaps the study involved a specific group of people, or the relation between the variables was only investigated in some situations.  Thus, it may be uncertain whether the correlational findings may generalize to other people or situations. 

Limitations of Randomized ExperimentsExperiments involving the random assignment of participants to conditions may allow us to make causal conclusions if the variables that are manipulated are not confounded with other variables.  However, there still may be limitations with respect to the generality of the findings.  The experiment may have involved a specific group of people, certain situations, and only some of the possible conceptualizations of variables.  Thus, we may not know whether the findings will be generalized to other people, situations, or conceptualizations of the variables.

Theoretical Framework LimitationConsider the variables self-esteem (belief in yourself) and self-efficacy (belief in your abilities). If the theoretical framework of the study was based on one of these variables, and the variable actually measured was the other, then the theoretical framework may be flawed. This limits the accuracy of the results. So a study is limited by whether the theoretical framework is an accurate reflection of the phenomena/variables they are studying.

The Phenomenon LimitationMany theoretical constructs measured in the social sciences are difficult to define. Review studies on concepts such as job satisfaction or student success, and different definitions will be found and different measurement instruments. The study is limited by the definition (how broadly or narrowly you define the phenomenon under investigation). For example, one might measure student success by test scores, graduation rates, or employment after completion. The results of the study will vary widely, depending on which of these definitions you select

The Instrument LimitationInstruments are limited by their reliability and validity. An instrument is reliable if it will give the same measurement every time when measuring the same construct. Consider weight, if you weigh yourself at home, then weigh yourself 5 minutes later, you should get the same measurement unless YOU have changed (put on/removed clothes). That's reliability.

13

Page 14: Assumption 2

If you weigh yourself at home, you may get a different measurement than at the doctor's office or the gym. Which scale is valid? If the density of the ground below each area is different, they ALL may be valid. If the density is the same, you need to determine WHICH scale is giving you the correct weight as only ONE is valid.

The Methodology LimitationThe results of the study are limited by the ability of the methodology to address the problem and purpose. The researcher might address the same research questions by several different methodologies. WHICH methodology you choose may increase or decrease the ability to find the answer being sought.

The Analysis LimitationThe results of the study will be limited by the ability of the statistical procedure selected to find statistical significance. The analysis must be appropriate to address the research question, and the test must have sufficient power to detect significance differences/relationships if they exist in the population.

The Power to Detect LimitationThe results of the study will be limited by the ability of the statistical test to detect significant differences/relationships if they exist in the population. It is likely not to find differences in the sample if they do not exist in the population.

The Participants LimitationThe study may also be limited by how well the participants in the study represent the population (this is why sample selection is so important! Nonrandom sampling means your participants likely do not represent the population). In addition, the results of your study are limited by the honesty of your participants, or their nonbiased participation (that is not changing their behaviour because they know they are participating in a study).

The Results LimitationFinally, the generalizability is limited by how well the sample represents the population. The study is also limited by the usefulness of the results to important stakeholders.Therefore the differences between assumptions and limitations are that while assumptions are factors potentially influential to your study for which you have no hard data, might not ever know, and can’t or don’t intend to control for, Examples may include: honesty or participant response, accuracy of proctored instruments, or utility of a particular instrument in the study (though not it’s validity and reliability this should be demonstrated through empirical support), the limitations are impediments to demonstrating the internal or external validity of the study, which should be relatively minor, as major limitations are sufficient basis to call for changing the study’s purpose, questions and/or methods. The examples may include: shortage or empirical data to support the hypotheses (though theoretical support should exist, or develop it), or the restrictions of descriptive, correlational or causal research.

14

Page 15: Assumption 2

BIBLIOGRAPHY

Mugenda, Olive M., and Abel G. Mugenda. 2003. Research Methods. Nairobi: Acts Press,

Kotheri C.R. 2004. Research Methodology. India New Age International (P) Ltd., Publishers.

Bouffard, M. (2000). Questioning philosophical and methodological research assumptions. Adapted Physical Activity Quarterly, 17(3), 340-358.

Brinberg, D.; & McGrath, J.E. (1985). Validity and the research process. Newbury Park, CA: Sage Publications.

Fain, J.A. (2004). Reading, understanding, and applying nursing research (2nd ed.). Philadelphia: F.A. Davis Company.

Fals-Borda, Orlando.  "Evolution and Convergence in Participatory Action-Research."  A World of Communities: Participatory Research Perspectives.  Ed. James Frideres.  Toronto: Captus University Publications, 1992.  14-19.

15