There is a very simple, straightforward test for paranormal abilities.
Two people, one who sends and one who receives. The sender thinks about something (e.g. a number, a color, a geometric shape), and the receiver reports what the sender is thinking. Repeat that a few dozen times. To make it all scientifically rigorous, you put them in separate rooms where they can't see each other. To make it even more rigorous, you require that they don't know each other, and possibly have never even met before the test.
That last requirement basically dooms the test to fail, at least with the way it would work to my personal understanding. But even without requiring that they don't know each other, the average person off the street still has no idea how to send or receive.
Most people broadcast their thoughts the way a light bulb gives off light. It goes out in all directions with no particular destination. It's stronger if there's any kind of emotional attachment to the thought, and weaker if the thought is neutral.
But to send a thought to a specific person, two things have to happen. First, you have to be able to find the other person. How close another mind is has nothing to do with physical geography, it has more to do with whether they're thinking about you (so it helps if they actually know you). How easy they are to identify depends on how well you know them - how well you can recognize their unique signatures. (Everyone has a unique signature.)
Also, sending a thought is more like beaming a laser. You want to focus it. Otherwise it takes a lot of energy and you get tired really really fast - and then you can't keep it up for the entire duration of the test.
Then there's receiving. Most people would think of it as sitting there passively, waiting for the outside thought to appear. However, a skilled receiver can also go over to the sender's mind and actively retrieve the thought (which is useful when the sender obviously has no idea how to send). This is easier if sender and receiver have compatible minds (are operating on the same wavelength, so to speak) - and, of course, if they are actually familiar with each other.
I once sat in on an informal run of this test at a college parapsychology club. The club leader sent, and everyone else wrote down what they thought he was thinking. We did numbers and colors, I believe. It took me all of the first round and part of the second to work out the mechanism involved (that is, the sender had no clue what he was doing and I had to go there) - so I scored low on the first one and high on the second. Statistically speaking, however, if your scientific experiment consists of aggregating everyone's scores in the room (which is what sensible scientists would do), your conclusion will be that it's not significantly different from random. Because most people don't know how to send or receive.