I wrote a research paper regarding the mental health of adolescents at a local magnet school, emphasizing the sources and affects of stress and the subsequent vulnerabilities it poses to developing forms psychopathology.

My problem: I need to formally establish that many students struggle with mental health problems. My initial idea was administering the BDI but, long story short, it can't happen. How else should I go about "formalizing" statistically the prevalence of psychopathology among the student body?

  • $\begingroup$ Maybe there are some official numbers available for comparable schools? $\endgroup$
    – K A
    Feb 13, 2016 at 2:48
  • $\begingroup$ Are there stats on certain high-risk behaviors, like being sexually active without proper protection, or use of heavier drugs? $\endgroup$ Feb 13, 2016 at 3:34
  • $\begingroup$ @RussellRichie I do not believe so. Would interviews be sufficient? $\endgroup$ Feb 13, 2016 at 3:41
  • $\begingroup$ @serk1 interesting suggestion, but unfortunately there is not. Especially since this magnet school is especially small (<200 students), logical comparisons are very hard to come by $\endgroup$ Feb 13, 2016 at 3:42
  • $\begingroup$ @cbutler16, you have interviews? Well dang, if you can get transcripts of them (are they transcribed yet? that's a mountain of work), then I am sure you can find correlates of psychopathology in them. I am sure there is plenty of lit on how behavior in such interviews correlates with psychopathology. $\endgroup$ Feb 13, 2016 at 3:57

1 Answer 1


The answer to this question depends on your ultimate goal. If you want to know whether or not students at this school are more likely to suffer from a psychopathology/s, you'll need to compare them to some other population. This other normative population can be students at another school, students in the country, students in the world etc. If you don't have access to data on another population you won't be able to use most of the truly inferential statisticsal procedures.

If you want to look at the school in isolation, you can use descriptive statistics and "describe" the situation at the school by looking at the way in which psychopathologies present, or, are distributed at this school. Additionally, you can have some arbitrary cut-point or score that you can use to make a judgment call. This probably won't cut it in an academic journal though, depending on the journal and field.

Reiterating Russell's suggestion, you can look at the correlation between the incidence of psychopathology and these stress factors you mentioned. This will at least allow you to compare these students to a hypothetical population of students where there is no correlation between psychopathology and these stress factors.

Your starting point will likely be to properly operationise your variables and then scouring the literature to see if you can glean something to compare or even make sense of your data from it.

First prize will be using an established and validated instrument. If you can't, you can still try and wing it, but be careful of the interviews though. To use the type of statistics I think you're hinting at, you need to properly operationilse your variables and quantify them. This is going to be hard to do with open ended questions, for example.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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