This is not a statistics course, but you need to conduct some analyses to help solidify your knowledge and understanding about statistical data analysis.  Now you will use the 'practice data.sav' file and SPSS to produce some descriptive and inferential statistics for the following questions.  Note that first you will need to decide what statistics to select - use the examples from above as well as your course text to help you determine the appropriate test.

Try to answer some of these questions for each other and post how you found the answers and/or strategies used - learn by sharing.  Please number each answer so your group members know which question you are responding to.  The questions for the Study Data File and Analysis assignment will be similar to those for discussion this week, so use your group discussion to help you understand what you need to do for your assignment.

1. Find two examples of data measures, other than the ones used above, where the level of measurement for the data is correct, and two where it is incorrect.  Note that virtually all the measures are reported correctly, but a few were changed to be inaccurate as an exercise for students to learn to recognize levels of measurement for data.

2. Why is it important to understand the level of measurement for data?

3. Choose a scale (at least interval) data variable other than 'age':

a) What is the most appropriate measure of central tendency?  (Also, interpret output).

b) How much does the sample vary?  (Also, interpret output).

4. Is there a statistically significant difference in marital status between male and female patients?

5. Is there a statistically significant difference in scores (means) for the Beck Anxiety Inventory (BAI) at baseline depending on patients' gender?

6. At baseline, does marital status for patients result in a statistically significant difference in BAI scores?

7. At the first data collection point, does the partners' experimental group have lower BAI scores than the partners' control group and is the result statistically significant?

8. What is the relationship between anxiety and uncertainty of illness for patients at time 3?

9. This question will be challenging to most of you, but try to answer anyhow.  Select appropriate statistical tests to test some of the study hypotheses - consider univariate and multivariate statistics:

a) Patients who participate in the post-discharge telephone intervention (treatment group) will have less uncertainty than patients who do not participate in the intervention (control group).

b) Regardless of study group, patients and their partners will have similar levels of quality of life at all data collection points.

c)  Partners' level of anxiety will decrease differentially over time in the treatment and control groups.