|
|
|
|
|
by plaxis
4780 days ago
|
|
No. Correlation is not causation. 1. There's no discussion of potential influencing variables and other inputs, like private tutoring and income that could affect scores. Also, note the sample: "The sample included 253 males and 67 females; 78% were Caucasian, 20% were Asian, and 2% were of other ethnicities." Why bother to add anything but white male at all? 2. since when do we take SAT at 13 Y/O?
3. Is the 1980's SAT even remotely similar to 2013? If not, so what?
4. "Top of their field" n=260 means we get sub samples of different industries around n= 3 to 16 or so (on graph). Please do 1 industry with n=31, and I might pay attention to a correlation.
5. Can we see the model output? Are the results significant, and the model sound? Based on the data, it would seem more likely that the influence on successes here is more likely that the individual is MALE and WHITE and NOT SAT score. I'm not sure I understand what this model is getting at except trying to suggest other than how to achieve material "success" through SAT SCORES??!! Read this one instead: http://www.businessinsider.com/grant-study-reveals-what-make... |
|