Also keep in mind that many students want to be someplace else too. It all really depends on your priorities as a student, which is never really made clear to applying undergraduates. If you are self-motivated and want to have the opportunity to get involved in a real laboratory and get experience as a research assistant, then go to an R1 university. You won't have as motivated instructors, but you'll have access to better facilities than you will elsewhere. But if you can't motivate yourself to learn from a book, and you need an instructor to tailor the material so that it's easier for you to learn it, you're probably better off at a small liberal arts college.

This is a little vague. The purpose of one of TFAs is to show how inaccurate the metadata on books in their database can be, and how Google is unwilling to do anything about it. Thus, when researchers use Google book search to look up information about books, rather than read the book (as the summary implies), they can be mislead.

Two examples from TFA: a search for "Internet" in books published before 1950 produces 527 results, and a book entitled "Culture and Society 1780-1950" was supposedly published in 1899.

They also defined effect size (a measure of the magnitude of a difference) as a measure of statistical significance (a decision based on the probability that an observed effect size would have occurred due to chance). But I suppose math/statistic confusion is even more common than experimental design confusion.

This is where RTFA comes in handy. The first paragraph of TFA:

But then what does this include? Isn't knowledge a psychological state? Affectivity? Willingness to communicate?
"You can't discriminate against me just because I don't know things."
"You can't discriminate against me just because I hate everyone around me."
"You can't discriminate against me just because I refuse to talk to my coworkers."
 
It's a slippery slope, which is why it hasn't been added already...

True - the correlation between intelligence tests and job performance are usually around .5-.6 and the correlation between conscientiousness (a personality trait) and JP is usually around .2-.3 - but it's better than nothing. But that's why I don't suggest small businesses use them. They only help organizational outcomes (profit) when used in the aggregate, over hundreds of applicants.

For reference, the correlation between unstructured interviews conducting by HR managers and JP is around 0...

Bah... nevermind. My math is terrible. :)

But you are still relying on the hiring manager having some degree of skill at identifying these people. And why is a test any more humiliating than an interview?

100 people + $10 per online test = $1000 expense

100 people x 15 minutes organizational time spent per interview x $20/h hiring manager = $30000 expense

Which would you go with?

Drug tests are actually related to job performance as well. If you correlate drug test pass/fail rates with supervisor ratings of job performance within companies that use drug tests but don't hire on them, there actually is a correlation - people who pass drug tests perform better. It's also easily legally defensible, which makes it a no-brainer for most hiring organizations.

Personality actually is correlated with job performance as well, and is job-related. The most empircally-supported personality trait is conscientiousness, which is a person's general tendency toward organization, orderliness, and meeting deadlines. Guess where that kind of tendency is useful. If you already like being organized, and your company wants you to be organized, you're probably going to be better at it than someone who doesn't like it.

Also, if you are conducting a major hiring effort to add 100 employees, your assertion probably is not true. 100 slightly above average employees will still better produce better overall organizational outcomes than 99 randomly hired employees and 1 exceptional one.