A.  Independent Samples t-test - Non-directional

For this exercise think of a research question, e.g. do mean scores on an anxiety measure differ by type of therapy group? You are asked to:

1. Formulate your own research question with corresponding null and alternate hypotheses.

2. Create your own data appropriate for your research question.

3. Test your hypotheses using the SPSS Independent Samples t-test procedure.

Begin constructing your data file by labeling your independent and dependent variables.  Designate variable types. Provide value labels for your independent variables (do you know why?).  Below is an example for the therapy group/anxiety score research question. DO NOT USE THE DATA BELOW, DOING SO WILL RESULT IN A 0 ON THIS ASSIGNMENT.

                THGROUP                           ANXSCORE

                1                                              30

                2                                              40

                1                                              50

                2                                              60

                1                                              25

                2                                              30

• Everyone should have different variables and data for this exercise, i.e., do not use the same numbers or variables associated with the example above, and don't use your friend's either. Create your own variables and data based on your research question.
• Provide data for 20participants, 10 for each level of your IV (think 20 rows in SPSS).
• At alpha = .05, the data that you create should yield a non-significant two-tailed p value (sig. on SPSS)associated with your t test. You may need to play around with the data to accomplish this.
• Make sure the homogeneity of variance assumption is not violated. Look in the Field text for further clarification on Levine's test.

Run the t-test analysis and then complete the following:

1. Clearly state your research question. In words, what are your null and alternate hypotheses (you will write the statistical hypotheses later in the step process).

2. Clearly state your independent variable (IV) and dependent variable (DV).

3. Write out the steps for this statistical problem (four steps) as explained in class and in your slides. Refer to your SPSS output for Step #4 (e.g. see output) - do not calculate the t test by hand.

4. What is the two-tailed significance level (sig or p value)? What does it mean within the context of your question?  Do you reject the null? Why or why not?

5. If your research question was directional and you needed to interpret the one-tailed significance level (p value), what isthe one-tailed pvalue/ sig level from your analysis?

6. Interpret the confidence intervals from your analysis. How do they provide you with information regarding whether to reject or retain the null?

7. Summarize your overall results clearly in paragraph form. Make sure to relate your results to your research question.

B. Paired/Dependent Samples t-test: Directional

For this exercise think of a research question, e.g. do mean scores on a spelling test improve after intensive study?  You are asked to:

1. Formulate your own research question with corresponding null and alternate hypotheses. 

2. Create your own data appropriate for your research question.

3. Test your hypotheses using the SPSS Paired/Dependent Samples t-test procedure. 

Here is an example of how to set up the data. Look in your SPSS text for more examples. DO NOT USE THE DATA BELOW; DOING SO WILL RESULT IN A 0 ON THIS ASSIGNMENT.

                                BE_STUDY                           AF_STUDY

                                 6                                             15          

                                 7                                             12

                                 8                                               5

                                 9                                             14

                                11                                           16

                                12                                           22

• As noted previously, everyone should have different variables and data for this exercise.
• Provide data for 20 participants (think 20 rows in SPSS).
• At alpha = .05, the data that you create should yield a significant one-tailed p value ("sig" on SPSS) associated with your t test. Keep modifying the data to accomplish this.

Run the appropriate t-test analysis and complete the following:

1. Clearly state your research question. In words, what are your null and alternate hypotheses?

2. Clearly state your independent variable (IV) and dependent variable (DV).

3. Write out the steps for this statistical problem (four steps) as explained in class and in your slides. Refer to your SPSS output for Step #4 (e.g. see output) - do not calculate the t test by hand.

4. What is the one-tailed significance level (p value)? What does it mean within the context of your question? Do you reject the null? Why or why not?

5. If your research question was non-directional and you needed to interpret the two-tailed significance level (p value), what is the two-tailed value from your analysis?

6. Interpret the confidence intervals from your analysis. How do they provide you with information regarding whether to reject or retain the null?  

7. Summarize your overall results clearly in paragraph form. Make sure to relate your results to your research question.

Need the following:

1. The responses to the questions at the end of each section.

2. The outputs from both parts A and B.