Hi, I need som second opinions on an issue.
I'm in the process of writing my thesis, where I want to explore the effect of social norms (SN) and social identity (SI) on engagement towards active breaks.
My supervisor is leading an RCT study with active breaks and helped me formulate my research question and hypotheses. In this, I want to see how SN and SI affect the engagement towards active breaks (e.g. the number of performed active breaks).
However, because the analysis will be based on a RCT-study I feel the effect of SN and SI is limited, as some people have been given tools to do active breaks and asked to do them at least 3 times a week, and some haven't. My research question fits more a 'natural' experiment, rather than a controlled one.
I feel it is hard to talk about the validity of my thesis, because there is such a mis-match between the research question and experimental design.
While my supervisor says it is not a problem, I am now starting to feel that I am blindly going down a path, that does not make sense.
(To my advantage, a lot of the participants did not comply with the study, in which I might be able to make an argument of the effect of SN and SI)
I guess I'm seeking some second opinions since my supervisor says it's fine, but my gut says something else.
Thanks in advance.
You gut is telling you have .indigestion. why is you research advisor your research advisor if you don't trust his opinion. Either do what he suggested or get another advisor. Your advisor is probably hoping for the second option because he is wondering why he agreed to take you since his gut is telling him that you are not going follow his advice so why should he waste his time on you. Good luck. David Booth
As I understand, the results of your study may be biased since the experiment does not go as it was initially planned. But, at the same time, you can conduct a simulation experiment and show/assess the potential bias. Having conducted this type of experiment will help i) understand the possible extent of the problem, ii) better communicate your concerns.
Valuable from a scientific point of view is both the confirmation of the given hypothesis and the negative result. But it's important not to hide the negative results. It is also not allowed to select the results that confirm the "only right truth" and ignore the others that do not confirm.
"Can you make valid research, when the experimental design does not match the research question" --- Ida Vendelø Jørgensen asks.
From your considerations, I got the impression that you are not feeling good because of the mismatch between your research question and the experimental design. I would feel the same if I were you. As I see it, when the experimental design does not match the research question, the research at hand, namely its findings, can raise problems of either internal or external validity. This seems not to be the opinion of your supervisor. May be you both need to discuss the issue you raise in an assertive manner. I take the liberty of suggesting you the reading of the following book : Achenbach, T. (1978). Research on developmental psychology: Concepts, strategies and methods. NY: The Free Press. It is a relatively old book, but it is, I think, a seminal book. Social norms (SN) and social identity (SI) are the two independent variables in your research. The USA psychologist Elliot Turiel has written a lot about societal norms or conventions- How do you define and operationalize the concept of SI?
I hope that this helps you, and good luck for your thesis Master or Ph D. thesis
The results of the research may be valid but will not answer the question, as the experimental design is different from the research question
I think you follow the supervisor and point the mismatch in your limitations of your study. I know you are scared, but since you know the study well I am sure you will convince your self. Research is like a journey. You ask how long the journey takes but along the line the car breaks down and you sleep on the road. This is still a journey so long as you explain it. FOLLOW THE LEADER FOR THIS ONE AND LEARN.
"Can you make valid research, when the experimental design does not match the research question" --- Ida Vendelø Jørgensen asks.
Have in mind that when a research is performed there two types of validity: Internal validity and external validity: Do have in mind only one of them or both? This may be important to answer your question.
BTw this student would just be using the collected data. Presumably his advisor is smart enough to know if that data fits the students research question. If you don't think the advisor can do that then you really need a new advisor don't you, Don't you? D Booth
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics