1.    Download the BMI Condensed SPSS data file 

2.   Open the data file within SPSS

3.   Select as interval-level dependent (outcome) variables respondent’s BMI (not grouped – BMI2) and TV watching hours per day. Calculate descriptive statistics and histograms for them (suppress the table for BMI). Note which measures are most meaningful, and interpret the results. 

4.   Perform a test of differences in mean TV watching and mean BMI by Gender and then by Race (Whites and Blacks only, which you can designate within the grouping variable specification), and then by one other variable of your choosing. Interpret the results of these six tests. 

5.   Perform a one-way analysis of variance of TV watching and BMI by Marital Status (in 3 groups), then by Race in four groups, and then by one other variable of your choosing, using Tukey’s HSD test as needed. Interpret the results of these six tests. 

6.    Perform a chi-square test of independence for the BMI group variable by one variable of your choosing (must be nominal or ordinal). Interpret the results of this test.

To run the chi-square test:

Use ANALYZE > DESCRIPTIVE STATISTICS > CROSSTABS

Enter BMI group variable in the rows and then your independent variable in the columns.

Go to Statistics and check off “Chi-square” and click CONTINUE

Go to Cells and check off “Column” under percentage.

  • Attachment 1
  • Attachment 2