Scribd
Upload
30 views

Program 1-Results

This document summarizes the results of a logistic regression model fit to Rasch model data with 20 observations and 2 response levels. It finds that the item effects are significant predictors of responses but the person effects are not. The model fits the data significantly better than the null model.

Uploaded by

Sherwin Balbuena
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
30 views

Program 1-Results

This document summarizes the results of a logistic regression model fit to Rasch model data with 20 observations and 2 response levels. It finds that the item effects are significant predictors of responses but the person effects are not. The model fits the data significantly better than the null model.

Uploaded by

Sherwin Balbuena
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Saturday, March 23, 2019 11:23:40 PM 1

The LOGISTIC Procedure

Model Information

Data Set WORK.RASCH

Response Variable score

Number of Response Levels 2

Weight Variable weight

Model binary logit

Optimization Technique Newton-Raphson

Number of Observations Read 20

Number of Observations Used 20

Sum of Weights Read 76

Sum of Weights Used 76

Response Profile

Ordered Total Total


Value score Frequency Weight

1 0 12 44.000000

2 1 8 32.000000

Probability modeled is score=0.

Class Level Information

Class Value Design Variables

item 1 1 0 0

2 0 1 0

3 0 0 1

4 -1 -1 -1

person 01 1 0 0 0

02 0 1 0 0

03 0 0 1 0

04 0 0 0 1

05 -1 -1 -1 -1

Model Convergence Status

Convergence criterion (ABSFCONV=0.0001) satisfied.


Saturday, March 23, 2019 11:23:40 PM 2

The LOGISTIC Procedure

Model Fit Statistics

Without With
Criterion Covariates Covariates

AIC 105.358 98.339

SC 105.358 105.309

-2 Log L 105.358 84.339

Testing Global Null Hypothesis: BETA=0

Test Chi-Square DF Pr > ChiSq

Likelihood Ratio 21.0196 7 0.0037

Score 19.1047 7 0.0079

Wald 15.6000 7 0.0290

Type 3 Analysis of Effects

Wald
Effect DF Chi-Square Pr > ChiSq

person 4 4.2081 0.3786

item 3 14.0768 0.0028

Analysis of Maximum Likelihood Estimates

Standard Wald
Parameter DF Estimate Error Chi-Square Pr > ChiSq

person 01 1 -0.3106 0.4835 0.4127 0.5206

person 02 1 0.4861 0.6796 0.5116 0.4744

person 03 1 -0.2541 0.4455 0.3252 0.5685

person 04 1 0.9170 0.5492 2.7878 0.0950

item 1 1 1.5513 0.5049 9.4424 0.0021

item 2 1 0.4478 0.4351 1.0588 0.3035

item 3 1 -1.2798 0.4865 6.9211 0.0085

Odds Ratio Estimates

Point 95% Wald


Effect Estimate Confidence Limits

person 01 vs 05 1.695 0.266 10.819

person 02 vs 05 3.761 0.373 37.906

person 03 vs 05 1.794 0.300 10.716

person 04 vs 05 5.786 0.783 42.729

item 1 vs 4 9.686 2.071 45.303


Saturday, March 23, 2019 11:23:40 PM 3

The LOGISTIC Procedure

Odds Ratio Estimates

Point 95% Wald


Effect Estimate Confidence Limits

item 2 vs 4 3.212 0.809 12.755

item 3 vs 4 0.571 0.130 2.504

Association of Predicted Probabilities and


Observed Responses

Percent Concordant 75.0 Somers' D 0.500

Percent Discordant 25.0 Gamma 0.500

Percent Tied 0.0 Tau-a 0.253

Pairs 96 c 0.750

You might also like