Class Exercise on Lagged Analysis
PM 536
December 3, 2020
Attached you will find output from the data collected to evaluate the ER Nutrition storyline that ran
last year. Answer the following questions:
1) How many respondents were lost to follow up and how many remained in the study?
2) Is there any attrition bias in the sample?
3) The dependent variable in this case is self reported behavior dichotomized on the mean
“srbehd”. Is self reported behavior at baseline (srbehd) correlated with self reported behavior
at followup?
4) What factors are associated with changes in self reported behavior?
5) What is wrong with this analysis?
6) We measured exposure to the ER storyline with a series of questions asking them how
frequently they watched ER and which of the episodes carrying our storyline they watched.
This scale is called “exposet2.” Was exposure associated with changes in self reported
behavior?