Exercise

Read the Green article Green Article. In your best prose answer the following questions:

1. What is the problem statement? Outline the rationale (logical argument) that this problem is worth studying. Is she consistent with the statement of the problem throughout the paper?
2. What are the independent and dependent variables in this study? Specifically, how are the operationalized?
3. To what group(s) can this work be generalized and why? (Threats to external validity?)
4. How was the instrument developed and what can you say about the quality of the instrument for this study? How does she justify the use of this instrument? Do you see any problems (threats to reliability or validity)? Take a close look at the questions in the instrument. Since the comparison data are not being gathered from a single instrument, what are the differences in the instruments and why?
5. How were the data collected and from whom? What can you say about the three different groups?
6. How were the data analyzed? Be specific here. You know a lot about doing statistical analysis so take a good look at her approach to analysis. Usually you do this by looking at the tables and seeing if you can understand them and then going back to the prose. Does the analysis she chose seem right based on things we have discussed in class?
7. What, specifically, did the author find? What did she present in the results section? Identify the main conclusions and the reasons for each. Does it appear that the conclusions are justified based on what was presented in the results section? Since she wrote out the hypotheses did she respond to all of them and what evidence did she use for that response?
8. Think about potential problems you have identified in this study. Does the author do a good job of outlining these in the conclusions section of the paper? (Limitations—Validity and reliability)
9. If you were studying this topic, what are a few things of importance that you could take from the reference list? Be specific. What does reading the reference list tell you about the topic and this study?
10. If you were including this paper in a literature review, what would you write?
    
    Don't let the strange analysis terms fool you. Almost all inferential statistics are based on the few principles we have already discussed and vary based on the relationship of independent and dependent variables. With that in mind, here are the terms you may not have seen before (you could look these up but I am doing it for you):
    
    Univariate Analysis (One Question at a Time)—Describes each variable in a set of variables individually. In cases where there is one independent variable and one dependent variable analysis is done with an ANOVA. Were there differences in the groups (independent) in how they answered each question (dependent)?
    
    Multivariate Analysis (Groups of Questions Simultaneously)—Analyzes multiple variables in a data set simultaneously. In cases where there is one independent variable and multiple dependent variables the statistics must account for the interaction among all of the dependent variables. Are there differences in how the groups (independent) answered on the accumulated score of all of the questions in the survey (dependent). Then a MANOVA (multivariate analysis of variance) is used. We haven’t studied how to do these but that is what Green did.
    
    Wilk’s lambda—Think of this as doing the same thing an ANOVA does (searching for group mean differences that were unlikely to ocurr by chance) but it does it when there are multiple dependent variables in the model (that is the MANOVA described above).