Hello together,
I am a master student from Germany and for the purpose of my thesis I made a model for the first time. The topic is "Privacy Concerns on Social Media". I am supposed to use SPSS and PLS for the analysis. However, one of my main dependent variables (privacy concerns) was not measured in my model with questions. I did not know that I have to use questions to measure it (long story and I guess misunderstandung). Now, I am desperately trying to find a solution. I already wrote a lot for my thesis, so changing the topic is not really an option. I explained the problem in a little bit more detail in the attached video. I would be really thankful if you share your opinion with me. Thank you
If you did not measure a variable, then there is no way that you can analyze it is any statistical program.
Do the independent variables measure privacy concerns or are they totally independent variables? If they are measures of the construct privacy concerns, then you can use first order (the independent variables) to create your second order construct (Privacy concerns).
We used Markland and Ingeldew's Exercise motivation theory in a paper and the first order constructs are used for the second order constructs. But they have to truly be second order constructs to use this method. In M&I, psychological factors such as stress management truly define psychological factors and we can use the items for the first order constructs to define the second order construct.
You can't do this if you are using self-efficacy, fear motivation, and numerous other variables to define privacy concerns.
See https://www.youtube.com/watch?v=kPeUTKjMF7o and many other articles on this type of construct to determine if your items work this way.
The solutions mentioned in the video have to be in accordance with what is considered in the theoretical framework or in the previous works. It is very risky for a thesis to move away much from the conceptualization, dimensions and scales established in the studied field.
Dear SG. Your question is important. I wonder why your supervisor did not guide you and check your work before starting the distribution of the questionnaire. However, it is too late now to fix the "questionnaire". I suggest that you run 5 or 6 focus groups and discuss with them 3 critical questions about this area. Read about focus groups or attend a workshop on how to plan and conduct focus groups. I trust this will help in fixing things. I suggest discussing this option with your supervisor and plan what you are going to do. Good luck. Prof S. Azer
- Badges
- Science topic
- Similar topics
- Social Science
- Media
- Social Media