Image from flicker.com
The research instruments used in a study must ultimately be fit for purpose – as the most appropriate to answer the questions asked, and no more.
For instance, a survey could be a suitable methodology for many purposes. They would also be efforts to have some degree of random sampling across school clusters and a reasonable sample size. However, as with all surveys, the generalisability of the findings could still be challenged. More importantly, the findings on the relationship between X and Y is co-relational at best; and not causal.
Given the nature of such findings, the use of the results must necessarily be circumscribed.
Such is the discipline in interpreting studies that must be adhered.