I am currently doing a research study about social media use and youth's well-being. My research questions are:

  1. How does the amount of time spent on social media affect the youth's well-being,
  2. How does the amount of social media following affect the youth's well-being,
  3. How does gender identity play a role in youth's well-being as a whole considering the
    (a) amount of social media following and
    (b) amount of time spent on social media, and
  4. How does grade level play a role in youth's well-being as a whole considering the
    (a) amount of social media following and
    (b) amount of time spent on social media.

I have already started my survey, and I can't do another survey again due to the limited resources and time that I have. I just want to know what statistical analysis / analyses should I use, using the data that I have?

  • $\begingroup$ Can you clarify the data that you have, & the structure of your study? You seem to have laid out your goals here, but we'll need more information to answer your questions. $\endgroup$ – gung - Reinstate Monica Sep 10 '17 at 4:18
  • $\begingroup$ The analysis is very straightforward, if you took a definitely representative sample of youths, randomly assigned them to the conditions you want to compare and objectively (at least with a very well validated instrument) directly assessed their well-being etc. before & after intervention. If you did not do any of these three things, this is an incredibly difficult question. Depending on which of these is not done that way starting over with a properly planned experiment might be best. $\endgroup$ – Björn Sep 10 '17 at 5:14
  • $\begingroup$ In my survey, I asked them what their gender and grade level are, how much time they spend on social media, how many social media following they have, and there was a rating scale that would measure their well-being (self-esteem, self-worth, and mental health (anxiety, depression, and hostility). I hope it's enough data to provide answers to the research questions. $\endgroup$ – pi3.14 Sep 10 '17 at 6:11
  • $\begingroup$ @Björn i'm sorry my research questions were wrongly crafted. I'm not intending to do an experiment for my study, I just to know the effect or correlation of the dependent variables to the independent variables. $\endgroup$ – pi3.14 Sep 10 '17 at 6:15

That's an obvious case for a regression with some continuous and some categorical predictors and some interaction terms.

It looks like you might have all the information you need, but it's obviously a very bad idea to start a survey without knowing how to analyze the data.

  • $\begingroup$ Yeah, I just have come to realize that. We immediately started with the survey because there was a deadline we had to meet. But can you please enlighten me more on how to do regression, because I am relatively new to using programs like stata and spss. $\endgroup$ – pi3.14 Sep 10 '17 at 2:03
  • $\begingroup$ @user75283 Using which program to analyze is not important at this step. You should first come up with your hypotheses and operationalize them. Then you should consider which programs to use. $\endgroup$ – Buomsoo Kim Sep 10 '17 at 2:08
  • $\begingroup$ How do i "operationalize" them? This is the first time that I'm doing a research like this. We weren't also taught about statistical analyses in our school, so I don't really know anything about stats (except for slovin's formula maybe). $\endgroup$ – pi3.14 Sep 10 '17 at 2:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.