I have a study in which I have one independent variable and three dependent variables with a moderating variable. During one of the discussions, someone told me that independent variables should be more than the dependent variables. I could not understand the reason behind it as one who was telling me failed to explain.
Can anyone please guide me if this is true and if it is true, what problems can occur in the study?
Hi Chaudry,
It depends on a lot more information than you provided in your question. For example, if the data were from a randomized controlled trial, then it is certainly possible to have more dependent variables than independent variables. In this case, it would be equivalent to running multiple t-tests with a continuous DV and a binary IV (representing the treatment variable). Of course, you'd need to run post-hoc tests to account for multiple testing.
On the other hand, if the data were from an observational study, then you would need to have several control variables in your model(s) to adjust for bias/confounding. In such a situation, we would expect that the number of independent variables would certainly outnumber the number of dependent variables.
I hope this helps
Ariel
It does not matter what is the number of variables in your study but what matters is that how you justify yourself in selecting your variables.
The theoretical approach and your model gives answer to your question.
Hypothetically, if you are studying the impact of one chemical/ spray on four type of crops and you are taking 1) water contents ( humidity), 2) temperature and 3) weather type as moderating variables. Then you are bound to take only 1 IV , 4 DVs and 3 Moderating variables. There is no fun to take additional IVs as it will be against your model under study.
Hope it is appealing to you!
Regards
Rafique
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Control Theory
- Control Systems