The goal of prevalence rates is to describe the health problem in the population; since this is the goal, the fact that it is only relevant to the population from which the data are taken is not a limitation.

Although theoretically, a genetic mutation could lead to a change in prevalence, this is a long-term process and not usually observed in abnormal research. 

Several factors may influence prevalence rates over time.  In addition, cultural, economic, and gender differences may account for differences in prevalence rates - so they are not universal.

Brown and Harris found that the most significant vulnerability factors were (1) Loss of one's mother before the age of 11, (2) lack of a confiding relationship, (3) more than three children under the age of 14 at home and (4) unemployment.

Although interviews were used to collect data, the study was a natural experiment.  There was a "before" and "after" television design.  The independent variable - the introduction of television - was not manipulated by the researcher.  It was an independent measures design.

An emic approach uses local knowledge to make a diagnosis. It is inductive in nature and attempts to be objective - that is, not influenced by the culture of the researcher.

Cheng (2018) found that education makes a difference.  However, the study was done also done on university students, so the generalizability of the findings may be limited.

As the study was trying to determine why more women are diagnosed with depression than men, it is not a problem that only women were studied - that was the population of interest. The sample size was 205, which is actually a reasonable size.  The fact that it was a quasi-experiment is not problematic as cause and effect were not really being investigated.  However, the study does not look directly at how clinicians make diagnoses.  Assumptions are made that the data provided by families to clinicians or reported to others may be biased.