The good news:
> There's been some evidence that the IAT is pretty robust. Most trivial matters like position of items don't much much of a difference. People who were asked to convincingly fake an IAT effect couldn't do it.
The bad news:
> A common critique of the test is that the same individual often gets two completely different scores taking the same test twice. As far as re-test reliability goes, .6 correlation is pretty good from a theoretical point of view, but more than enough to be frequently embarrassing. It must be admitted: this test, while giving consistent results for populations, is of less use for individuals wondering how much bias they personally have.
Otherwise, I don't think the implicit bias is what OP said. Our HR would remind us recency bias, for instance, during a perf review. That kind of implicit bias does exist and is worth reminding.
Teasing out those differences could help e.g. layout information and design cameras and image pipelines to reduce the effects of bias.
These tests have been administered to large numbers of people and on average, almost every single person that has taken the test has scored some level of implicit bias. As a result, it's very likely (but not certain) that you ARE unaware of your implicit bias.
Of course, if you take the tests and score perfectly, you'll now be able to demonstrate empirically that you have no measurable implicit bias and will have an answer to those people who insist you do.
The reason why this is different to the walking on water statement, is that there are hundreds of thousands of data points all showing implicit bias is almost universal, whereas there are zero data points showing people can walk on water after believing in Jesus.