4 ms·
A survey on a self-selected group is kind of worthless. Eg, their results show a plurality of libertarians, which may be the case in reality but is much more li
by mtraven 9y ago
A survey on a self-selected group is kind of worthless. Eg, their results show a plurality of libertarians, which may be the case in reality but is much more likely to be an effect of libertarians being more likely to take the time to fill out the form.
- bkohlmann 9y agoThat's a valid criticism. However self-selected group surveys can show important trends. For instance, if they said that "as a whole, 25% of Silicon Valley believes that they are discriminated against," that could clearly be an invalid interpretation. However, if they said, "150 people who identified as libertarians took this survey, and among them, 45% felt like their productivity was inhibited because of perceived viewpoint bias, whereas of the 100 self-identified liberals who took it, only 10% felt the same," that would be a very telling statement. [these numbers are made up for illustrative purposes only]