# The inkblot test - what is it for?
**URL:** https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101
**Category:** Factual Questions
**Created:** [October 17, 2003, 9:28pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101 "2003-10-17T21:28:57Z")
**Posts on this page:** 20
**Page:** 1
### Author:  [@Kinthalis](https://boards.straightdope.com/u/Kinthalis)
#### Post date: [October 17, 2003, 9:28pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/1 "2003-10-17T21:28:57Z")
we’ve always seen psychologists in TV and movies subject their patients to those inkblot tests.
What I want to know is, what are these tests for? What do they say about a person’s psychological health? Are there any right or wrong answers? And finally, what are the resuylts of such a test good for? Can a doctor in a court of law say “Well he saw a bunny where there was none this proves he is insane!” ???
---
### Author:  [@Eats\_Crayons](https://boards.straightdope.com/u/Eats_Crayons)
#### Post date: [October 17, 2003, 11:00pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/2 "2003-10-17T23:00:34Z")
The “ink blot test” is the Rorschach Test. The one’s you see in movies and TV are supposed to be fake because the real blots are supposed to be kept “secret” from the general public so that your reactions to them are immediate and are not prejudiced by what you may already know about them. However, a few publications, both in print and [on-line](http://www.deltabravo.net/custody/rorschach.htm) will show you the outlines.
Some psychologists feel the test is remarkably unreliable, others swear by it.
No response is often a sign of neurosis, if you keep seeing human genitalia they’ll also flag your file. Most “healthy” folks see things like butterflies, the silhouette of a fox, two people dancing etc.
---
### Author:  [@Eats\_Crayons](https://boards.straightdope.com/u/Eats_Crayons)
#### Post date: [October 17, 2003, 11:03pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/3 "2003-10-17T23:03:12Z")
Now back to my Dope moratorium…
---
### Author:  [@TheLoadedDog](https://boards.straightdope.com/u/TheLoadedDog)
#### Post date: [October 17, 2003, 11:09pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/4 "2003-10-17T23:09:35Z")
“Und vot do you think of ven I show you dis vun, mmm?”
“SEX!”
“Interesting. Und zis vun?”
“Oh, that’d definitely be sex!”
“I see. Un zis last vun here. Vot is the first thing that…”
“SEX!”
“I am seeink a pattern formink here. You are obsessed vith sex, und I think…”
“_I’m_ obsessed? YOU’RE the one with the dirty pictures!”
😃
---
### Author:  [@ianzin](https://boards.straightdope.com/u/ianzin)
#### Post date: [October 17, 2003, 11:25pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/5 "2003-10-17T23:25:02Z")
It’s a crock of horse spit, basically. A book has recently been written about it, and there was a good article in a recent edition of Skeptical Inquirer. For more info:
[http://www.csicop.org/si/2003-07/rorschach.html](http://www.csicop.org/si/2003-07/rorschach.html)
---
### Author:  [@markci](https://boards.straightdope.com/u/markci)
#### Post date: [October 18, 2003, 12:23am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/6 "2003-10-18T00:23:22Z")
It’s NOT a crock of shit. I’ve had one. I think the accuracy may have a lot to do with who is doing the interpretation – there are different theories on how to score and interpret it. Mine was done by a guy who is considered one of the leading Rorshach experts.
The test can tell a lot about your personality and mental health. I received a computer-generated printout – a pretty lengthy one describing personality traits in detail. Nearly all of them matched traits I had identified myself, or which had come out in therapy, prior to taking the test.
The statements about me were specific – not fortune-cookie statements that could apply to anyone. For example, the test indicated introversion, which is certainly the case. If it had indicated extroversion, which it just as easily could have, I would have laughed at it.
And it was administered by a technician who’d never met me, and scored by a computer, so it wasn’t a case of the scorer skewing the results because he knew me.
Administered by the right person, I’d say it definitely has some validity.
---
### Author:  [@Ringo](https://boards.straightdope.com/u/Ringo)
#### Post date: [October 18, 2003, 12:49am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/7 "2003-10-18T00:49:52Z")
The [Skeptic’s Dictionary](http://skepdic.com/inkblot.html) has an entry that includes:
> [@](#):
>
> Rorschach testing is inherently problematic. For one thing, to be truly projective the inkblots must be considered ambiguous and without structure by the therapist. Hence, the therapist must not make reference to the inkblot in interpreting the patient’s responses or else the therapist’s projection would have to be taken into account by an independent party. Then the third person would have to be interpreted by a fourth ad infinitum. Thus, the therapist must interpret the patient’s interpretation without reference to what is being interpreted. Clearly, the inkblot becomes superfluous. You might as well have the patient interpret spots on the wall or stains on the floor.
And:
> [@](#):
>
> To avoid this logical problem of having a standard for a standard for a standard, etc., the experts invented standardized interpretations of interpretations. Both form and content are standardized.
A friend had to take one as part of a child custody battle. One comment said friend made was that the ink “blots” all have bilateral symmetry, common in biology; i.e., of course you see moths, people and sex organs. Heh, I don’t know, I’ve never taken one.
But they sound about as reliable as polygraph tests, i.e., crock of horse spit. IMHO, of course.
---
### Author:  [@markci](https://boards.straightdope.com/u/markci)
#### Post date: [October 18, 2003, 1:44am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/8 "2003-10-18T01:44:59Z")
Based, of course, on no knowledge or experience whatsoever.
---
### Author:  [@Ringo](https://boards.straightdope.com/u/Ringo)
#### Post date: [October 18, 2003, 2:18am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/9 "2003-10-18T02:18:55Z")
> [@](#):
>
> Based, of course, on no knowledge or experience whatsoever.
Eh? What, you took the test and were impressed by the results? Why did that make you feel it was a valid test?
---
### Author:  [@Achernar](https://boards.straightdope.com/u/Achernar)
#### Post date: [October 18, 2003, 4:27am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/10 "2003-10-18T04:27:29Z")
> [@](#):
>
> \*Originally posted by Ringo \*
> A friend had to take one as part of a child custody battle. One comment said friend made was that the ink “blots” all have bilateral symmetry, common in biology; i.e., of course you see moths, people and sex organs. Heh, I don’t know, I’ve never taken one.
But bilateral symmetry is common in a lot of things. Cars, computers, baskets, planets, lamps, airplanes, and books, off the top of my head.
I think you need to give psychologists a little credit. Yes, if something is ingrained in tradition, it’s hard to drop once it’s been debunked, but if something had good, solid evidence against it, like horoscopes, psychologists wouldn’t be using it.
My intro psych textbook cites a couple of studies on this. They say that _individual_ Rorschach indices are not very valid by themselves (Zubin, Eron, and Shumer, 1965). However, when you compare assessments made by Rorschach experts based on _overall_ results, and those made by psychiatric evaluators, they show a mean correlation of +0.21 (Little and Shneidman, 1959).
---
### Author:  [@ianzin](https://boards.straightdope.com/u/ianzin)
#### Post date: [October 18, 2003, 5:28am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/11 "2003-10-18T05:28:18Z")
> [@](#):
>
> \*Originally posted by markci \*
> \*\*Based, of course, on no knowledge or experience whatsoever. \*\*
In my previous post I linked to an article which in turn refers to a book that has been written about this very question - the validity of the Rorschach test. Read the article, or read the article in the Skep Inquirer, or read the book. They will provide all the ‘knowledge or experience’ you could ask for.
To suggest that something ‘seems’ accurate based on your individual subjective, personal evaluation of ‘accuracy’ of a single instance is perfectly acceptable as a reported opinion; it is absolutely a non-starter as a way of assessing whether something works. To take an extreme and hypothetical example, suppose I put a piece of paper saying ‘Your name is markci’ inside every one of 1000 numbered envelopes, and asked you to choose one based on the last three digits of your phone number. You’d be amazed at the ‘accuracy’, and the surprising correlation between phone numbers and the information in my envelopes. Of course, the other 999 times that I do this on 999 other people, they can see it’s rubbish.
This is an extreme and purely hypothetical instance, but it serves to illustrate an important point about subjective evaluations based on one personal experience. How do you know the Rorschach test didn’t, so to speak, ‘get lucky’ with you, but would be hopelessly inaccurate at other times? How do you know they don’t detect ‘introvertism’ in every single punter?
For more on the dangers of subjective evaluation of ‘accuracy’, go to Google and search on ‘cold reading’. It could be quite an eye-opener for you. I’ve done cold reading demos for TV. I’ve had people rate my readings as 99.5% accurate, and say they were highly specific, when in fact the so-called ‘accurate’ readings are just instances of cold reading and I know nothing whatsoever about the person I’m giving the reading to. Subjective evaluation of single instance accuracy is just notoriously unreliable as a way of discovering anything much. Read the article, read the book, look at the evidence. The Rorschach test is a crock of horse spit.
---
### Author:  [@scr4](https://boards.straightdope.com/u/scr4)
#### Post date: [October 18, 2003, 5:37am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/12 "2003-10-18T05:37:43Z")
> [@](#):
>
> \*Originally posted by Eats\_Crayons \*
> \*\*The one’s you see in movies and TV are supposed to be fake because the real blots are supposed to be kept “secret” from the general public so that your reactions to them are immediate and are not prejudiced by what you may already know about them. \*\*
They’re also copyrighted, which I think is the main reason they’re not reproduced. Psychology/psychiatry tests are pretty expensive too.
---
### Author:  [@susan](https://boards.straightdope.com/u/susan)
#### Post date: [October 18, 2003, 5:55am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/13 "2003-10-18T05:55:38Z")
There are different opinions among psychologists about the Rorschach’s reliability and validity. This is true of all tests, whether of intelligence, psychopathology, or personality. With all due respect to the Skeptic’s Dictionary, though, it is not necessary that the therapist view the blots as “without structure.” This opinion may be based on older ways of scoring the Rorschach, which considered it a “projective” instrument. These days, most psychologists tend to think of it as a way to understand the client’s style of perception and thinking.
As do many psychologists, I use a scoring system that’s called “Exner scoring” and I find the Rorschach very useful at times. I would caution anyone taking this test against advice or descriptions such as “No response is often a sign of neurosis, if you keep seeing human genitalia they’ll also flag your file,” above. This is an inaccurate representation of the test that is perpetuated in cartoons and television. Like the MMPI and many other instruments, many aspects of Exner scoring have been validated with different populations, with normal response ranges determined empirically.
If you don’t believe the Rorschach is valid and don’t want to take it, tell the psychologist this. If you don’t think it’s valid for custody determination, again, tell the psychologist this and request an alternative. I’ve never “subjected” a “patient” to it. This kind of language implies that psychologists are not acting for you, but against you. This may be true in some limited forensic settings, but is not true of psychologists in general. Psychologists in general have the intention to be helpful.
---
### Author:  [@PhilAlex](https://boards.straightdope.com/u/PhilAlex)
#### Post date: [October 18, 2003, 5:59am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/14 "2003-10-18T05:59:34Z")
ONe of the books by William Poundstone, I thiink Bigger Secrets had a good chapter on this, too.
---
### Author:  [@Cardinal](https://boards.straightdope.com/u/Cardinal)
#### Post date: [October 18, 2003, 10:59am UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/15 "2003-10-18T10:59:56Z")
My college friend from a farm took one from the doctoral psych student in the dorm for practice, and he came up with an amazing number of mammals. I never would have seen those, so I thought it had some kind of point.
---
### Author:  [@BlackKnight](https://boards.straightdope.com/u/BlackKnight)
#### Post date: [October 18, 2003, 3:38pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/16 "2003-10-18T15:38:16Z")
The May 2001 issue of Scientific American ran an article on the subject, written by the authors of the CSICOP article already linked to above.
Just throwing that datum into the mix.
---
### Author:  [@Ringo](https://boards.straightdope.com/u/Ringo)
#### Post date: [October 18, 2003, 4:04pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/17 "2003-10-18T16:04:22Z")
> [@](#):
>
> \*Originally posted by Achernar \*
> However, when you compare assessments made by Rorschach experts based on _overall_ results, and those made by psychiatric evaluators, they show a mean correlation of +0.21 (Little and Shneidman, 1959).
What does that mean?
---
### Author:  [@Achernar](https://boards.straightdope.com/u/Achernar)
#### Post date: [October 18, 2003, 5:20pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/18 "2003-10-18T17:20:13Z")
Just to make sure I’m not misinterpreting what I’m reading, here’s the exact quote (_Basic Psychology_, 5th ed., p. 562):
> [@](#):
>
> Studies of this kind have sometimes been criticized as too reductionistic, for the focus on a single aspects of an examinee’s Rorschach record. Wouldn’t it be better to use the test as a whole, allow the judge to read the entire record verbatim (or even to administer the test), and then predict the criterion on the basis of this overall, global knowledge? One study that meets these conditions used twelve eminent Rorschach experts who tried to assess the personalities of various patients on the basis of their complete Rorschach records. These Rorschach-based assessments were then compared to the pooled judgment of a number of psychiatrists who had read each patient’s case history, obtained in six or so interviews of sevelar hours each. The mean correlation between the Rorschach experts’ predictions and the psychiatrists’ judgments was +.21. Apparently then, global assessment on the basis of the verbatim record does have some modest validity (Little and Shneidman, 1959).
So what this means, I think, is that they got the Rorschach experts to read the responses to the Rorschach test, and asked them a yes-or-no question. They got psychiatrists to read the case history, and asked them the same question. The two agreed about 60.5% of the time, whereas flipping a coin would agree 50% of the time. (I know that not every yes-or-no question has a 50% chance of being yes, but they have statistical methods of correcting for this.)
---
### Author:  [@Ringo](https://boards.straightdope.com/u/Ringo)
#### Post date: [October 18, 2003, 6:00pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/19 "2003-10-18T18:00:01Z")
It’s been about 25 years since I took a stat course, but isn’t +.21 a pretty _low_ positive correlation? I’m not arguing, just asking. Enlighten me.
---
### Author:  [@Achernar](https://boards.straightdope.com/u/Achernar)
#### Post date: [October 18, 2003, 6:21pm UTC](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101/20 "2003-10-18T18:21:58Z")
Yes, it is low, which is why they call it modest. But it is high enough that you can count on it being real, that is, not just a statistical anomaly.
[Next page](https://boards.straightdope.com/t/the-inkblot-test-what-is-it-for/208101.md?page=2)