Tender-Heartedness and Attitude About Animal Research

First Things First

·  Go here and download to your computer (not open in your browser) the Power Point slide for creating the diagram.

·  Download the data using the “SEM Tenderheartedness” link on my SPSS Data Page. The data are in the form of a variance/covariance matrix. If you are going to use SAS Calis instead of Amos, download ARC-VarCov.dat from my StatData Page.

Description of the Data

Four of the items on Forsyth’s Ethics Position Questionnaire seem to measure “tender-heartedness.” The items are:

·  Risks to another should never be tolerated, irrespective of how small the risks might be.

·  It is never necessary to sacrifice the welfare of others.

·  If an action could harm an innocent other, then it should not be done.

·  The existence of potential harm to others is always wrong, irrespective of the benefits to be gained.

These items were included in the idealism subscale employed in the research reported in:

  • Wuensch, K. L., & Poteat, G. M. (1998). Evaluating the morality of animal research: Effects of ethical ideology, gender, and purpose. Journal of Social Behavior and Personality, 13, 139-150.

The subjects were also asked to answer a question about how justified a particular case of animal research was (higher scores = more justified) and asked whether or not that research should be allowed to continue (0 = no, 1 = yes).

Your Assignment

Your assignment is to use the data gathered in this research to test the SEM model diagrammed below. It includes two latent variables: Tender Heartedness (high scores = high tender heartedness) and Animal Research (high scores = favor animal research). Obtain standardized estimates, squared multiple correlations, and estimates of variance.

In a Word document, enter your answers to each of the following questions. Immediately following each answer, paste in the relevant part of the text output from Amos or SAS.

  1. Is the Chi-square test (of the null that the fit is as good as that of the saturated model -- perfect) significant?
  2. Do any of the regression weights fall short of statistical significance?
  3. Which three paths have the highest absolute standardized regression weights?
  4. Interpret the association between Tender Heartedness and Animal Research and specify the proportion of the variance in Animal Research that is explained by Tender Heartedness?
  5. Which of the indicators for Tender Heartedness has the highest reliability?
  6. What is the relationship between the estimated reliabilities and the standardized weights of the paths leading to the indicators from the latent variables?
  7. Does the GFI indicate that the model fits the data well?
  8. Does the RMSEA indicate that the model fits the data well?
  9. You know that the Chi-square test will be significant even when the fit is good if you have a sufficiently large sample. How large would the sample need be here for the Chi-square to be significant at the .05 level?

Rather than using the diagram created by Amos, I would like you to use a diagram I drew in PowerPoint. Open it in PowerPoint Click on each “vari” (in a text box) and replace “vari” with the estimated variance of the error or disturbance. Click on each “.rr” and enter the estimated reliability of the indicator variable. Click on each “sw” and enter the estimated standardized weight. After you have finished entering the parameter estimates and reliabilities, File, Save as, and select a graphics format (png, jpg, or gif). Then open your Word doc and Insert, Picture the diagram into the Word document below your answers to the eight questions posed above.

Print the Word document and bring it to me in class on Wednesday the 18th of November, 2009.

Karl L. Wuensch, November, 2009