Rorschach, Research and Wikipedia

08/31/2009 05:12 am ET | Updated Nov 17, 2011

There was a bit of a tussle in the world of psychology recently when The New York Times published an article discussing the controversy of the publication of the Rorschach inkblots (yes, the inkblots made famous in the psychological test, the Rorschach Inkblot Test) on Wikipedia. This despite the fact that the inkblots themselves have been available on Wikipedia for well over a year.

How odd that anyone would make such a fuss over a relatively arcane, projective psychological test, and certainly not over one that is as old as the Rorschach and rarely used outside of a full psychological testing battery. There exist a number of readily-available books that talk about the interpretation of the Rorschach, available on or even in your local university library. If anybody wanted that badly to see what the infamous inkblots looked like, they could easily already do so.

What Wikipedia has done, then, is to simply make the viewing much easier, bringing it into the public spotlight.

On one side you have psychologists who decry, "But if people can review our psychological instruments ahead of time -- before they have to take such a test for real -- then it will affect the results." That very well may be the case, but there's very little evidence that actually shows that any type of pre-exposure to the inkblots taints the results. In fact, I couldn't find a single study that examined this issue. So while it is a widely-held belief amongst many psychologists, it's not a belief grounded in scientific data for the Rorschach (what some psychologists might refer to as an "irrational belief").

On another side, you have people who make the claim that none of this matters much anyway, because the Rorschach is one of those old projective tests that isn't really very valid or scientific. This is a criticism I myself once believed as well, being schooled in the more modern, "objective" scientific instruments like the Minnesota Multiphasic Personality Inventory-2 (MMPI-2). But a reading of the overall literature on the Rorschach paints a more nuanced picture.

The claim that the Rorschach isn't really valid was recently once again forwarded by Wray Herbert. Herbert is not a psychology professional, researcher or clinician, but rather is the director of public affairs for the Association for Psychological Science, writing today on the issue over at Newsweek:

But this heated debate has failed to raise (or answer) the most important question of all: does the Rorschach work? The answer is no, and here is the best evidence.

That was a pretty emphatic and clear-cut answer -- NO! -- the Rorschach is a piece of sh**. But, as usual, the real answer is a little less black and white, and far more subtle. It's hard to paint a complex picture with a single, broad brush stroke. In fact, the authors of the paper he cites remind readers as much:

At the same time, dismissing the Rorschach in broad brush as invalid oversimplifies the genuine state of affairs. Meta-analyses of published research on the Rorschach suggest that at least some Rorschach indexes possess above-zero validity...

In other words, yes, the Rorschach does work for some things. Which Herbert admits further down in the article:

With the exception of schizophrenia and similarly severe thought disorders, the Rorschach fails to spot any common mental illnesses accurately. The list of what it fails to diagnose includes depression, anxiety disorders, psychopathic personality, and violent and criminal tendencies. It also can't detect sexual abuse in children, even though it's used for that purpose.

In fact, Lilienfeld et al. (2000) admitted quite a bit more, finding that the Rorschach can correctly identify schizophrenia, borderline personality disorder, schizotypal personality disorder and bipolar disorder. It also seemed to be able to help clinicians identify thought disturbance, psychotherapy prognosis (wow, that's something valuable to have!), and dependency.

All of this is assuming that the Rorschach is typically used in some sort of assessment vacuum, where only its data alone is used to make judgments about a person. In reality, nothing could be further from the truth. In modern practice, the Rorschach is most typically administered along with a a half-dozen or more other psychological instruments, including other projective and objective tests (like the TAT, MMPI-2, an IQ test, etc.). And these are rarely or routinely administered to most people seeking help for a mental health issue. They are administered when there is a serious question as to the person's personality or cognitive functioning that can be gleaned from a simple diagnostic clinical interview.

Because it's administered in a larger set of psychological tests, the psychologist who analyzes the test data is looking for trends or similarities amongst the disparate tests. A single odd score on the Rorschach would likely be dismissed, if not confirmed by other psychological testing data.

A year after Lilienfeld et al. published their research in 2000, Irving Weiner from the University of South Florida published a study in a competing journal from the American Psychological Association in 2001 that called into question many of the criticisms of Rorschach detractors. In a later research article, Weiner (2005) rebuffed the findings of the 2000 Lilienfeld et al. study altogether:

Rorschach critics have nevertheless argued that the validity data for the Rorschach are not sufficient to warrant its use (Lilienfeld, Wood, & Garb, 2000). If true, this assertion in light of the Hiller et al. (1999) findings would indicate that neither the MMPI nor any other currently available personality assessment instrument should be used by practicing psychologists. These critics have maintained that the Hiller et al. meta-analysis is flawed, which seems unlikely given the care with which it was conducted and the methodological sophistication of those who conducted it. These same critics have pointed out that the RIM does not correlate well with the MMPI, while ignoring not only method differences between the instruments but also research showing that the RIM and MMPI correlate quite well when people respond to both instruments in an open and forthcoming manner, as opposed to being guarded and defensive on either or both.

Which brings up another observation -- that battles over tests like the Rorschach are epic within psychology. Professionals, psychologists and researchers all seem to take sides, and then argue their side to the death. In reviewing the research literature for this article, you can't help but notice the clearly delineated "sides" here, and notice the same researchers constantly calling into question the Rorschach, and another set of researchers constantly defending it. Such divisions often take on the character of a religious battle, one where scientific 'truth' is just another tragic victim to logical fallacies, biased data interpretations and egos.

The truth of the Rorschach lies somewhere in-between -- it is a useful psychological instrument for the thousands of psychologists who use it to help them figure out the psychological makeup of people in need. Its science, while not ideal, is actually more robust than many professionals know. And if people want to take a look at the Rorschach cards over at Wikipedia, they are welcomed to do so (likely without causing serious harm to the profession of psychology).


Hiller, J. B., Rosenthal, R., Bornstein, R. F., Berry, D. T. R., & Brunell-Neuleib, S. (1999). A comparative meta-analysis of Rorschach and MMPI validity. Psychological Assessment, 11, 278-296.

Lilienfeld, S.O., Wood, J.M. & Garb, H.N. (2000). The scientific status of projective techniques. Psychological Science in the Public Interest, 1(2), 27-66.

Weiner, I.B. (2005). Rorschach Assessment in Child Custody Cases. Journal of Child Custody, Vol 2(3), 99-119.

Weiner, I.B. (2001). Advancing the science of psychological assessment: The Rorschach Inkblot Method as exemplar. Psychological Assessment, 13(4), 423-432.