https://implicit.harvard.edu/implicit/Study?tid=-1Done it before, not impressed. Too many other possible explanations for the results (as pointed out in the questions at the end).
There are none so blind as those who will not see.The topic of this task was randomly assigned to you from a list of dozens of topics. You are welcome to complete as many sessions as you wish, and every session will be a different topic. Some will be topics you have thought about many times, others might be new or unusual topics that you have not considered before. Just return to the login page and enter your email address to start again
Thus confirming exactly what I said...I just took the one about weapons. It says that I associate weapons slightly more with blacks than with whites. Since I'm part black, does that make me racist against myself?
The problem is that the test concluded I associate weapons more with blacks than whites, which is not true. The result had nothing to do with subconcious association and everything to with the "weapons/white" catagory being presented first and run through twice. Obviously the testee will do poorly with this association because they're unfamiliar with the test. After practicing the game twice, and getting the hang of the multiple association, they then present you with the "weapons/black" catagory. You can now go through this exercise quickly and accurately because of the previous practice, not because you associate blacks with weapons. They then conclude that you associate weapons with blacks based on that result - not true. Start the test over, now that you're practiced, and I guarantee your results are different.No1 padawan has it just right, correlation is not cause and I doubt 'StyleZ' has wits enough to criticize Implicit Attitudes Testing' protocol.
As I said, it is interesting, it tests an attitude toward weapons [obligatory On Topic] and is no more valuable than a shoot/no-shoot exercise.
Believe nothing you read or hear without verifying it yourself unless it fits your preexisting worldview.