Im conducting a study where I want to investigate consumer attitudes and purchase intention via questionnaires. I know that product related attitudes (price, taste etc) affects purchase intention, but I'm interested in investigating other areas. Do I need to include these product related attitudes in my questionnaire and control for them in my analysis, or can i disregard/exclude these variables entirely from my study?
Thanks in advance.
Controlling for variables improves a study’s internal validity by reducing the impact of confounding variables or other external variables. Therefore, all variables that potentially influence the outcomes should be controlled—apart from the independent and dependent variables. In observational studies, controlling for variables can be done using statistics. Here is an interesting read.
Pearl, J., & Mackenzie, D. (2018). The book of why: The new science of cause and effect. Basic Books. https://www.basicbooks.com/titles/judea-pearl/the-book-of-why/9780465097616/
Good luck,
Yes, please include all those variables that are not in your model but could potentially influence the variables in your model, especially the outcome variables. Good luck!
We already know that control variables seek to rule out alternative answers to those sought in the study. They are factors that influence the dependent and independent variables but are not the object of the study. The differentiating aspect of control variables is that they are considered alien to the model, unrelated to the hypotheses being tested (SPECTOR; BRANNICK, 2011). However, many studies include control variables with no specific theoretical interest, sometimes without a clear justification for their use or present reduced comments. Given this context, I suggest pondering whether there is a theoretical and methodological assumption that justifies the need for the inclusion of these variables.
- Badges
- Science method
- Similar topics
- Philosophy