|
|
Hello,
I have a question regarding some analyses.
Following are the objects of our study:
1. To assess consumers' drug-related knowledge across different ad
formats (six) and determine the format that produces more knowledgeable
consumers.
2. To determine the factors affecting consumers' drug-related
knowledge.
To address the first objective I conducted one-way ANOVA with knowledge
being the dependent variable. The F-test was significant with p-value =
0.0009.
Since we were interested in comparing the (five) new brief summary formats
with No brief summary format (control), Dunnett's test was uses for post-hoc
analyses. The results indicate that the three new formats -- QA, bulleted
list and nutritional facts panel produced significantly higher knowledge
scores than No brief summary formats. Furthermore, pairwise t-tests
confirmed that there was no significant difference in knowledge scores among
QA, bulleted list and nutritional facts panel formats.
My question is:
I want to find out the factors that have an impact on consumers'
drug-related knowledge. For this purpose I was thinking of conducting a
multiple regression. I am kind of confused as to which approach/analysis I
should use in order to answer this question. It is because here we have
three groups.
Kindly give me some hints/suggestions as to how should handle this
situation.
Thank you,
Nilesh
|
Locked
