I took a class in testing while at the University of West Florida. A few things surprised me.
One was that true/false tests are very reliable, and multiple choice are also among the best.
This was back in the 60s but at that time IQ tests were among the most reliable of all. In other words if you took another one with different questions you likely would score a very similar score.
To be fair, the situation I mentioned was kind of unusual.
Test 1 was a more or less “standard” IQ test. Scored 132.
Test 2 had the person administering the test read me a long essay on an obscure subject, then answer multiple choice questions on the topic. The topic in this case was ancient Greek religious practice, cult and priests, as opposed to mythology. Scored 169.
Problem (for accuracy of second test) was I’d read a long work on the subject just a couple of weeks before. Wasn’t particularly obscure to me. :)