One thing to look at with these tests is the percentage inclination toward a specific trait. For instance, if someone is 70% inclined toward introversion, they are very much set in that trait and unlikely to show a different result depending on the day. However, if they are only 5% inclined toward Perceiving, then the answers they have given to the P/J questions have been mixed, and the are only slightly tilted in favour of the P. That would mean that if they did the test again (or another similar test) and answered one or two of those questions differently (because they felt differently on the day) then the result may come out different. So as an example - the first time I did the test my results were - Introverted: 74%, Intuitive: 39%, Feeling: 5%, Judging: 20% I've done the same test or other tests probably 10 times since. 8/10 times I get INFJ, which means that's a pretty solid pattern. But on a couple of occasions, because the Thinking/Feeling and to a lesser degree Judging/Perceiving percentages are not so pronounced, I get a slightly different result. And interestingly, when I read the description of the different profile, I see parts of myself in that too. So what does it all mean? Well, these tests break down 6 billion people into 16 basic personality profiles (which is hilarious when you think about it). But that doesn't mean that everyone is equally inclined to all the traits within the profile. It just gives an indication of how our personality is inclined - what our innate tendencies are. For ones who are getting different results every time they do the test, it's probably an indication of a good balance between one side of a trait and another. Neither is dominant. See http://www.16personalities.com/articles/strength-of-individual-traits and http://www.humanmetrics.com/cgi-win/JungType.htm Anyway, its all interesting and can teach us a bit about how we are 'wired', which can be very helpful. But like all social science, its based on human wisdom, and as such is not exact.