multiple classification analysis (mca) using winidams (widyo pura buana)

28
12/7/2012 Widyo Pura Buana (MCA using WinIDAMS) 1 MULTIPLE CLASSIFICATION ANALYSIS (MCA) USING WinIDAMS WIDYO PURA BUANA Widyo Pura Buana (MCA using WinIDAMS) Jalankan Software WinIDAMS Widyo Pura Buana (MCA using WinIDAMS)

Upload: prawiraningbumi

Post on 28-Oct-2014

81 views

Category:

Documents


7 download

DESCRIPTION

Multiple Classification Analysis digunakan untuk melihat pola hubungan antara dependen variable (interval) dan independent variable (ordinal/nominal). Pengolahan data dapat menggunakan Excel, SPSS atau WinIDAMS.

TRANSCRIPT

Page 1: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 1

MULTIPLE CLASSIFICATION ANALYSIS(MCA)

USING WinIDAMS

WIDYO PURA BUANA

Widyo Pura Buana (MCA using WinIDAMS)

Jalankan Software WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Page 2: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 2

Tampilan Awal WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Create an Application Environment (1)(Membuat/Menentukan Tempat Kerja)

Widyo Pura Buana (MCA using WinIDAMS)

Page 3: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 3

Create an Application Environment (2)(Membuat/Menentukan Tempat Kerja)

Widyo Pura Buana (MCA using WinIDAMS)

Create an Application Environment (3)(Membuat/Menentukan Tempat Kerja)

Widyo Pura Buana (MCA using WinIDAMS)

Page 4: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 4

Kembali ke Windows Semula

Widyo Pura Buana (MCA using WinIDAMS)

File Data dalam Format Excel(Akan di Import ke dalamWinIDAMS)

Widyo Pura Buana (MCA using WinIDAMS)

Page 5: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 5

File XLSX  Save ke CSV Format

Widyo Pura Buana (MCA using WinIDAMS)

Import ke WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Page 6: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 6

Pilih File yang akan diimport keWinIDAMS (1)

Widyo Pura Buana (MCA using WinIDAMS)

Pilih File yang akan diimport keWinIDAMS (2)

Widyo Pura Buana (MCA using WinIDAMS)

Page 7: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 7

Pilih File yang akan diimport keWinIDAMS (3)

Widyo Pura Buana (MCA using WinIDAMS)

Hasil Import ke WinIDAMS (1)

Widyo Pura Buana (MCA using WinIDAMS)

Page 8: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 8

Hasil Import ke WinIDAMS (2)

Widyo Pura Buana (MCA using WinIDAMS)

Hasil Import ke WinIDAMS (3)

Widyo Pura Buana (MCA using WinIDAMS)

Page 9: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 9

Variable Definition

Widyo Pura Buana (MCA using WinIDAMS)

Build Dictionary

Widyo Pura Buana (MCA using WinIDAMS)

Page 10: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 10

Save Dictionary Data

Widyo Pura Buana (MCA using WinIDAMS)

V2, V4, V7 dan V21 sudah masuk

Widyo Pura Buana (MCA using WinIDAMS)

Page 11: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 11

V2 dirubah menjadi

Widyo Pura Buana (MCA using WinIDAMS)

Pendefinisian Variabel (1)

Widyo Pura Buana (MCA using WinIDAMS)

Page 12: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 12

Pendefinisian Variabel (2)

Widyo Pura Buana (MCA using WinIDAMS)

Pendefinisian Variabel (3)

Widyo Pura Buana (MCA using WinIDAMS)

Page 13: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 13

Pendefinisian Variabel (4)

Widyo Pura Buana (MCA using WinIDAMS)

Membuat File Setup (SYNTAX)

Widyo Pura Buana (MCA using WinIDAMS)

Page 14: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 14

Window Setup/Syntax

Widyo Pura Buana (MCA using WinIDAMS)

Syntax MCA ‐WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Page 15: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 15

OUTPUT MCA WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Output MCA WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Page 16: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

12/7/2012

Widyo Pura Buana (MCA using WinIDAMS) 16

WIDYO PURA BUANAMULTIPLE CLASSIFICATION ANALYSIS

USING WinIDAMS

Widyo Pura Buana (MCA using WinIDAMS)

Page 17: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 1 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

*** UNESCO WinIDAMS 1.3 December 2007 *** 7/12/2012

9:27:11

____________________________________________________________________

| |

| Welcome to WinIDAMS 1.3 December 2007 |

| (English Version) |

| |

|____________________________________________________________________|

1 $COMMENT SETUP FILE WITH EXAMPLES

2 $COMMENT OF 1 DATA MANAGEMENT

3 $COMMENT ANOVA AND MULTIPLE CLASSIFICATION ANALYSIS

Listing of setup

1 $RUN MCA

2 $FILES

3 DICTIN = 'MCA WinIDAMS.dic'

>>>>> E:\MCA_IDAMS\data\MCA WinIDAMS.dic

4 DATAIN = 'MCA WinIDAMS.dat'

>>>>> E:\MCA_IDAMS\data\MCA WinIDAMS.dat

7 $SETUP

8 'Analysis of variance and multiple classification analysis'

9 BADDATA=MD1

10 DEPVAR=V4 CONVARS=(V1,V2,V3)

11 DEPVAR=V4 CONVARS=V3

Page 18: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 2 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

**** MCA **** Multiple Classification Analysis

Source: OSIRIS III.2 (MCA), Univ. of Michigan, U.S.A.

Last updated: UNESCO, 25 August 2005

No main filter

Label: Analysis of variance and multiple classification analysis

Parameters:

BADDATA=MD1

Parameters as interpreted:

Input ddname suffix: IN

All cases (after filtering) will be used from the input file

Bad data will be replaced by MD1

Analysis specifications:

DEPVAR=V4 CONVARS=(V1,V2,V3)

DEPVAR=V4 CONVARS=V3

***W* MCA001 --- Multivariate predictors must be in the range 0-31

After filtering, 101 cases read from the input data file

Page 19: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 3 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Analysis - 1 Analysis of variance and multiple classification analysis

Dependent variable

Name #ARTICLES *

Number 4

Max.code 9999999.000

Include MD1? NO

Include MD2? NO

Include outliers? YES

Range: L.T. -26.519

G.T. 35.487

Weight variable? NO

Print frequencies? NO

Iteration maximum 25

Convergence test PCTMEAN

Test for convergence 0.00500

Print coefficients? NO

Number of predictors 3

Predictor list

Variable Name Number of codes

1 CM POSITION IN UNIT * 3

2 SEX * 4

3 SCIENTIFIC DEGREE * 6

Number of cases eliminated

due to dependent variable requirements: 8

due to weight requirements: 0

due to predictor requirements: 0

Number of cases remaining: 93

Number of outlying cases: 0

Page 20: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 4 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Results based on test 2 Iteration 14

Dependent variable statistics

Dependent variable (y) = 4: #ARTICLES

M e a n = 4.4838710

Standard deviation = 6.2006536

Coeff. of variation = 138.3

Sum of Y = 417.00000

Sum of Y square = 5407.0000

Total sum of squares = 3537.2258

Explained sum of square = 1173.0288

Residual sum of squares = 2364.1970

Number of cases = 93

Page 21: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 5 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Predictor statistics

Predictor: Variable 1 CM POSITION IN UNIT

Unadjusted No of Sum of Per Class deviation from

Class Label cases weights cents mean grand mean Coefficient Adjusted mean Stand dev. C. var.

1 HEAD 8 8 8.6 9.12500 4.6411290 -4.7486949 -0.26482391 4.6425824 50.9

2 S&E 67 67 72.0 5.13433 0.65045738 1.3889226 5.8727937 6.5410024 127.4

3 TS 18 18 19.4 0.00000 -4.4838710 -3.0593460 1.4245250 0.0000000 *******

Eta-square = 0.15903983 Beta-square = 0.13516912

Eta = 0.39879799 Beta = 0.36765352

Eta-square(adj)= 0.14035182

Eta(adj)= 0.37463558

Unadjusted deviation SS = 562.55975

Adjusted deviation SS = 478.12369

Predictor: Variable 2 SEX

Unadjusted

No of Sum of Per Class deviation from

Class Label cases weights cents mean grand mean Coefficient Adjusted mean Stand dev. C. var.

1 MALE 28 28 30.1 5.92857 1.4447002 0.48761001 4.9714808 6.9704592 117.6

2 FEMALE 63 63 67.7 3.93651 -0.54736304 -0.20506388 4.2788072 5.8554442 148.7

4 1 1 1.1 0.00000 -4.4838710 0.43825150E-01 4.5276961 0.0000000 *******

5 1 1 1.1 3.00000 -1.4838710 -0.77786005 3.7060108 0.0000000 0.0

The correction of Eta-squared for the number of subclasses was too large. Value of

eta-squared is 0.28164053E-01

Eta-square = 0.28164053E-01 Beta-square = 0.28026458E-02

Eta = 0.16782150 Beta = 0.52940022E-01

Eta-square(adj)= 0.0000000

Eta(adj)= 0.0000000

Unadjusted deviation SS = 99.622620

Adjusted deviation SS = 9.9135914

Page 22: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 6 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Predictor statistics

Predictor: Variable 3 SCIENTIFIC DEGREE

Unadjusted

No of Sum of Per Class deviation from

Class Label cases weights cents mean grand mean Coefficient Adjusted mean Stand dev. C. var.

1 PROFESS. 6 6 6.5 11.0000 6.5161290 10.777214 15.261086 3.2863353 29.9

2 ASS.PROF 6 6 6.5 13.0000 8.5161285 8.2397604 12.723631 8.4380092 64.9

3 DOCTOR 25 25 26.9 5.00000 0.51612902 -0.23260523 4.2512655 6.6833126 133.7

4 M.A/M.SC 42 42 45.2 3.52381 -0.96006155 -2.0998745 2.3839965 5.0327614 142.8

6 OTHER 13 13 14.0 0.00000 -4.4838710 -1.4517369 3.0321341 0.0000000 *******

9 MD 1 1 1.1 0.00000 -4.4838710 -1.2194612 3.2644098 0.0000000 *******

Eta-square = 0.28744265 Beta-square = 0.37308550

Eta = 0.53613681 Beta = 0.61080724

Eta-square(adj)= 0.24649109

Eta(adj)= 0.49647868

Unadjusted deviation SS = 1016.7496

Adjusted deviation SS = 1319.6876

Page 23: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 7 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Multiple classification analysis statistics

R-squared (unadjusted) = proportion of variation explained by fitted model = 0.33162

Adjustment for degrees of freedom = 1.12195

***Multiple R (adjusted) = 0.50011 Multiple R-squared (adjusted) = 0.25011

Page 24: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 8 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Dependent variable 4: #ARTICLES

Listing of Betas in descending order

Rank Var. no. Name Beta

1 3 SCIENTIFIC DEGREE 0.61080724

2 1 CM POSITION IN UNIT 0.36765352

3 2 SEX 0.52940022E-01

***Multiple R (adjusted) = 0.50011 Multiple R-squared (adjusted) = 0.25011

Page 25: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 9 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Analysis - 2 Analysis of variance and multiple classification analysis

Dependent variable

Name #ARTICLES *

Number 4

Max.code 9999999.000

Include MD1? NO

Include MD2? NO

Include outliers? YES

Range: L.T. -26.519

G.T. 35.487

Weight variable? NO

Number of predictors 1

Predictor list

Variable Name Maximum code

3 SCIENTIFIC DEGREE * 9

One-way analysis of variance only

Number of cases eliminated

due to dependent variable requirements: 8

due to weight requirements: 0

due to predictor requirements: 0

Number of cases remaining: 93

Number of outlying cases: 0

Page 26: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 10 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

Predictor category statistics

Code Label N Weight-sum % Mean S.D.(est.) C. var. Sum of Y % Sum of Y-square

1 PROFESS. 6 6 6.52 11.0000 3.28634 29.9 66.0000 15.83 780.000

2 ASS.PROF 6 6 6.52 13.0000 8.43801 64.9 78.0000 18.71 1370.00

3 DOCTOR 25 25 27.17 5.00000 6.68331 133.7 125.000 29.98 1697.00

4 M.A/M.SC 42 42 45.65 3.52381 5.03276 142.8 148.000 35.49 1560.00

6 OTHER 13 13 14.13 0.00000 0.00000 ******* 0.00000 0.00 0.00000

9 MD 1 1 0.00000 0.00000 ******* 0.00000 0.00000

Totals 92 92.00 100.0 4.53261 6.216693 137.2 417.0000 100.0 5407.000

***Note: all these statistics exclude codes with fewer than two cases

Page 27: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 11 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

One-way analysis of variance statistics

***Note: all these statistics exclude codes with fewer than two cases

Eta-squared (unadjusted) = proportion of variation explained for 5 catagories = 0.28332

Adjustment for degrees of freedom = 1.04598

***Eta (adjusted) = 0.50037 Eta-squared (adjusted) = 0.25037

Total sum of squares = 3516.9021

Between means sum of squares = 996.42627

Within groups sum of squares = 2520.4758

F( 4, 87) = 8.598

***** Normal termination of MCA

***** No more RUN statements in SETUP; step terminated

Page 28: Multiple Classification Analysis (MCA) Using WinIDAMS (Widyo Pura Buana)

| 12 Widyo Pura Buana Pustaka ( MCA Menggunakan WinIDAMS)

SYNTAX (FILE SETUP WinIDAMS)

$COMMENT Setup file with examples

$COMMENT of 1 data management

$COMMENT ANOVA and multiple classification analysis

$RUN MCA

$FILES

DICTIN = 'MCA WinIDAMS.dic'

DATAIN = 'MCA WinIDAMS.dat'

$SETUP

'Analysis of variance and multiple classification analysis'

BADDATA=MD1

DEPVAR=v4 CONVARS=(v1,v2,v3)

DEPVAR=v4 CONVARS=v3

DATA

No V2 V4 V7 V21 No V2 V4 V7 V21 No V2 V4 V7 V21

1 1 1 1 12 36 2 2 3 2 71 3 2 9 0

2 2 2 3 3 37 2 2 4 5 72 3 2 6 99

3 2 2 4 1 38 2 2 3 6 73 1 1 1 11

4 2 1 3 2 39 2 2 4 3 74 2 2 2 22

5 2 2 3 15 40 2 2 4 2 75 2 2 4 2

6 2 2 4 2 41 2 2 4 2 76 2 1 4 0

7 2 2 4 3 42 2 1 2 23 77 1 1 1 12

8 2 2 4 0 43 2 1 4 5 78 2 1 3 4

9 2 1 4 21 44 2 2 4 18 79 2 1 3 4

10 2 2 4 2 45 2 2 4 0 80 2 2 4 4

11 2 2 4 2 46 2 2 4 5 81 2 2 4 1

12 2 2 4 5 47 2 2 3 4 82 2 2 4 1

13 2 2 4 1 48 2 1 4 0 83 3 2 6 99

14 2 5 4 3 49 2 2 3 1 84 3 1 6 99

15 2 2 4 0 50 3 2 6 0 85 1 2 2 6

16 2 2 4 3 51 3 2 6 0 86 2 2 3 2

17 2 2 3 0 52 3 2 6 0 87 2 2 3 2

18 2 2 2 11 53 1 1 1 5 88 2 1 4 1

19 3 2 6 0 54 2 2 3 17 89 3 1 6 0

20 3 4 6 0 55 2 2 3 1 90 1 1 1 11

21 3 2 6 0 56 2 2 3 3 91 2 1 4 1

22 1 2 3 1 57 2 2 4 2 92 2 1 4 1

23 2 2 2 2 58 2 1 4 6 93 2 2 2 14

24 2 2 3 25 59 2 1 4 1 94 2 2 4 0

25 3 2 3 0 60 2 2 4 5 95 2 2 4 0

26 3 2 3 0 61 2 1 4 6 96 3 1 6 0

27 3 2 4 0 62 2 2 4 10 97 3 1 6 0

28 3 2 4 0 63 2 2 3 3 98 3 1 6 99

29 3 2 6 0 64 2 1 3 2 99 3 1 6 99

30 3 2 6 0 65 2 2 3 4 100 3 2 6 99

31 3 1 6 0 66 2 2 3 2 101 3 2 6 99

32 3 2 6 0 67 2 1 3 20

33 1 1 1 15 68 2 1 3 2

34 2 2 4 20 69 2 1 4 1

35 2 2 4 3 70 3 2 9 99