Letters to the Editor
-
Desigirl, what would the control have been?
Desigirl, my understanding is that this was a study designed to see if there is a difference between two population samples.
The study doesn't seem to make any predictions that there would be a difference, or explain the difference.
It is merely analyzing the populations to see if there is a difference.
There was no independent variable being manipulated in an experiment.
What would the control have been? People that didn't take the test (and so were not measured?)
Having found a difference, the next step might be to hypothesize reasons for that, and design an experiment to test
those hypotheses. That experiment could have control and experimental groups, people that either were not, or were manipulated in some manner, perhaps by having them take a course, or play tetris, or change their gender, or possibly sit next to their partner, or apply makeup, or dye their hair blonde, or paint their nails, or read some blogs on how to properly blame the patriarchy.
You can question if the test is a valid predictor of anything, and you can question how they obtained their sample. I will say a sample of almost 200,000 people appears to be nothing to sneeze at, but I too am leary about internet data collection. They do cite others regarding that method of collection.
If anyone can think how or why this study required a control, I would be grateful. It's been too long since my intro to science courses...