By Michael A. RoBards, LCSW
You took one test and got INFP. Another said INFJ. A third, on a bad day, said ISFP. Now you have three labels and less clarity than when you started. This is one of the most common frustrations people bring to typing communities, and the fix is not a fourth test.
Every MBTI-style questionnaire measures four continuous traits and then forces each one into a binary. If your real position on Thinking–Feeling is close to the middle, whether you land on the T or the F side depends on how the questions are worded and how you felt that morning. Different tests also draw their cutoff lines in slightly different places. So three tests disagreeing does not mean one of them is right and two are wrong. It usually means you are near the middle on at least one dimension, and the tests are doing exactly what they are built to do.
That is worth taking in: a near-the-middle result is information about you, not a failure of the test or of your self-knowledge.
Here is what I walk people through in my office. Instead of answering questions about yourself in the abstract, look at what you actually do, across months, not on a Tuesday.
If your behavior says you are between INFJ and INFP, read both full profiles slowly. You will almost always recognize yourself more clearly in one, and the places where you see yourself in both are simply true about you. The profiles on this site are written for exactly this: INFJ, INFP, and the rest of the 16 types.
One more thing that helps more than people expect: ask a partner, sibling, or close friend which description sounds like you. We are experts on our own inner life and often blind to how we actually show up. The people around you see your extraverted side better than you do.
Do not retake the same test hoping for a tiebreaker. A 51/49 split will just flip the coin again. And do not type yourself as the version you would like to be. The most useful type is the one that explains the patterns you already have.
The Insight Game is built on the clinical Type and Temperament work of Dr. Martine J. RoBards, Ph.D. (1946–2017). Live community accuracy data is on the research page.