Rorschach: Men Who Stare at Blots
Please accurately describe the featured image:
A) It’s the Eiffel tower with two crabs waving flags at it!
B) It’s my dreams of being a circus clown lying broken on the floor!
C) It’s just some ink smeared around, in card ten of the Rorschach test.
If you picked C, you’d be a winner! (The prize is +1 internet points and a big internet hug.) Rorschach tests are inkblots. Ten of them, each printed on a card. They do not come with instructions. There are no instructions. The test is an adaptation of a parlor game. The test is a projective test, meant to diagnose pathology and personality, from the dustbin of psychology. It’s been tossed away with penis envy and homosexuality-as-a-mental-disorder. Or so I thought. Turns out, Rorschach tests are still being administered. Let’s talk skeptically about that.
Herman Rorschach created the inkblot test in the 1920’s, with the standard deck of 10 cards. That was the only standardized thing about it. Patients were handed one card after another, and expected to describe what they saw. Based on the description, you could be diagnosed. Seem fishy?
Others agreed, and fifty years later, John Exner created a standard system for analyzing responses to the Rorschach cards. Seriously, for fifty years, everyone was basically inventing their own way to interpret their clients. Exner’s system involves rating responses on 100 different criteria. Did the client look at the whole inkblot? Did he/she see what everyone else did? Was the color of the blots discussed? Then all of these assessments get turned into a profile, that spits out a diagnostic result.
Unfortunately, Exner’s method fails the two tests of a useful diagnostic measure: reliability and validity.
Reliability means that if Doc John and Doc Jill both administer the test to Patient Lee, they should each come up with the same diagnosis. In the same vein, if Patient Lee stares at inkblots in 2011 and again in 2012, he should get the same results. Doesn’t happen.
To be valid, the Rorschach test has to do what it says it does: accurately say who has mental illness and who doesn’t. Again, no dice. Like most projective tests, it’s likely to give you a false positive–telling you your brain is wonky when it’s fine–called overpathologizing. This is really bad–nothing like wandering around and paying money to treat something you don’t have.
So why, WHY are we still administering Rorschach tests? Why are we training new students how to use them? Eighty percent of surveyed clinical psychologists administer them at least occasionally. Forty two percent give them regularly or always. They give these tests on their clients dime, when the entire establishment knows they don’t give the same results across time or practitioners, don’t actually measure mental illness, and tend to tell healthy people that they have a disorder. Why has this continued? Why isn’t anyone throwing fits?
As biological and neurological research has taken off, and testing has become increasingly standardized, diagnosis has appeared to become something anyone with training in correctly scoring a test can do. We’re getting closer to being able to spot mental illness via the brain and brain waves. Projective tests are something the psychology establishment has a corner on–because so much is (wrongly) left up to the interpretation of the administrator, it’s something that feels like it cannot be taken away–even if all the evidence says it should be.
Universities keep teaching their grad students to administer the Rorschach test, because it’s another line on their resumes. If the old establishment is still using it, the next generation of students will keep learning it. And we will keep describing the puppies, pigs, people, and pelvises we see in hundred year old ink spots. Let’s end that. The Rorschach test does not work, and there’s real harm and monetary loss if we keep pretending it does.
Other discredited projective tests you should ignore: Szondi Test, playing with anatomically correct dolls to diagnose sexual abuse, graphology, the Hand Test, and the Luscher Color Test.
Featured images via Wikimedia Commons
Credit to: Lilienfeld, S.O, Wood, J.M., & Garb, H.N. (2001) Picture This. Scientific American.