3 ms·
One thing that jumps at me in this study is the high risk of investigating two different populations with different characteristics. Foreign students are more l
by zmk_ 11y ago
One thing that jumps at me in this study is the high risk of investigating two different populations with different characteristics. Foreign students are more likely to be high achievers (and high-achieving women), quantitative degrees are also in higher demand from foreigners and more likely to be in English. If you are Dutch you can pick whatever university/field you want and they will admit you (exception is medicine); thus there is a higher chance of low-achievers (who drive the results) to be in that part of the sample. I am suspicious of results which are driven by the tail of the distribution and die out in time (when that tail drops out).
- tomp 11y agoDidn't the study compare the performance of a group before and after the marihuana ban, with the natives being the control group? the differences you list (motivation, skill, effort) didn't change (except as an effect of the ban).
- losvedir 11y agoThe study used a "diffs-in-diffs" technique which would control for that. It takes the difference between two groups, say English and Dutch, prior to banning marijuana for English students, and then looks at the difference between the two groups afterwards. So it's a way of looking at how the performance of the English group changed with banning the drug, while trying to control for the things that affected _all_ students.