I have a research design where I have a 3x2 mixed ANOVA, focussing on attitudes towards a drug. The between participants part has 3 levels, with 2 different methods of communicating about risks of taking the drug, and one control, where there is no explanation about the drug risks. The within particpants part has 2 scenarios, in which they imagine they have 2 different sets of symptoms that they have gone to the doctor about, but the doctor will recommened against the use of the drug. I then want to measure 3 different DVs, one is attitudes towards the drug (on a scale), trust in their doctor (also a scale) and for each scenario I will have a yes/no on whether they will take the drug or not. I am planning on running 2 ANOVAs for the two seperate continous DVs, but I am unsure how to statistically analyse whether the yes/no is significant (I want to know which interventions/types are they more likely to go against dr advice, and if the information provided is informative enough for them to make the best decision). So what tests should I do?
1
-
1$\begingroup$ I'd recommend consulting with a statistician. Logistic regression is an option but that is a modeling approach, not a "test". Similarly "ANOVA" is an analysis approach, not a design. $\endgroup$– Bryan Krause ♦Commented Jan 6 at 22:16
Add a comment
|