Tag Archives: Cambridge Face Memory Test

Radio show on prosopagnosia from earlier this year

What’s in a face? Prosopagnosia. All in the Mind. ABC Radio National. January 21st 2018.




Super-recognizers on Australian public radio today


Genelle Weule So, you think you’re good at recognising faces. ABC Science. March 11 2018.


Super-recognisers. All in the Mind. ABC Radio National. 11 March 2018.


Leading researchers and prosopagnosics on Australian radio show about the extremes of face memory ability

Malcolm, Lynne What’s in a face? Prosopagnosia. All in the Mind. Radio National. February 19th 2017.



Dean, Diane Prosopagnosia: What it’s like to live with ‘face blindness’. ABC News. February 20th 2017.



There’s a link to the Cambridge Face Memory Test at the webpage for the radio show.

Sad but not surprising that prosopagnosics can be mistakenly perceived as having a personality issue rather than a perception issue.

I’ve got an anecdote that is rather like the opposite of the one shared by Dr Karl in the radio show, in which he came to understand that he had an unusual problem in recognizing fellow-students at university. Just this week I recognized (with no foreknowledge) a student that I once shared a tutorial group with when the student made a very brief appearance as an actor in a television advertisement. Nice! I wish him the best of luck in his acting career. He was working in a series of health promotion ads that consistently feature better acting than that often seen on the TV shows.

In case you are wondering, the music used in the radio show were two hits of the 70’s; that famous tune by Grace Jones and at the end the big hit by Roberta Flack.

The search for an effective intervention for prosopagnosia continues, but at least the knowledge of what is going on must be some help to people who face this challenge.



Recent articles about supers, prosopagnosia, policing and face recognition research

Keefe, Patrick Radden The detectives who never forget a face. New Yorker. August 22nd 2016. Print edition title: Total Recall.


I was glad to read in this substantial and interesting article that face identification was not the only evidence used to convict criminals. And the last couple of sentences in this piece are too true!


Montagne, Renee ‘New Yorker’: The Detectives Who Never Forget A Face. NPR. August 17th 2016.



Large twin study using the CFMT reportedly finds face recognition is heritable but largely independent of general intelligence and object recognition ability



I wish I had the full scientific background to fully interpret this interesting new study, because the results have HUGE implications in psychology, but as far as I know are not particularly surprising or at odds with related research. The genetic and phenotypic independence of face recognition ability would smash to smithereens the long-debated idea of “g”, or one (mysterious) factor largely determining general mental ability. Face recognition or face memory appears to defy “g”, but all the same, I can’t help clinging to the idea that there’s a link between top ability in face recognition and at least some other cognitive gifts. Based on personal experience I find it hard to leave behind the idea of a link between elite reading and writing ability, synaesthesia and superior face recognition.

Placing the heritability of face recognition ability at 61%, as this study has done, kicks sand in the face of the long and bitterly debated idea that giftedness or talent is the result of long hours of focused training rather than innate ability, but I can think of one researcher who has championed the “trained not innate” position on talent or expertise for many years, who seems to lack an awareness of the entire body of face recognition research, instead focusing his attentions on elite performers in sport, music, memory and chess. Ignorance is bliss, they say.

I am a super-recognizer, and I have no memory of ever training my ability in recognizing or memorizing faces, and no one has coached, pressured nor trained me to this specific task. I defy those who argue that intelligence is “environment” not genetics to explain me and faces. Up until a few years ago I had no idea I was even above average with faces, so don’t ask me.

Supermatchers super-recognizers, same thing, isn’t it?

Passport Problem. Catalyst. ABC. February 24th 2015. http://www.abc.net.au/catalyst/stories/4185916.htm

Why are the science journalists at the Catalyst team trying to distance this story from the existing body of research and writing on the same subject as the story, by using a new term for people with tested elite ability in face recognition? As far as I can tell, the skill is the pretty-much same as the skill measured by the CFMT, which I believe is recognized by researchers around the world as a gold-standard test of face memory or face recognition.

Calling all supers! New testing opportunity for all, and for some to take part in a study


One of the world’s leading researchers in the related areas of face memory, face recognition, prosopagnosia and super-recognition has given me the tip that super-recognizers are wanted as research subjects, and they are being recruited through the above web link which appears to be associated with the long-running online research and volunteer testing website TestMyBrain. I have also been advised that for this study the researchers are looking for subjects who reside in the continental states of the United States of America, but if that isn’t you, and you get a very high score in the test and follow the instructions and send the researchers your score there is some possibility that some time in the future you might be sought for some other study. Of course, you don’t have to share your score with anyone, and you might wish to take this test simply to get a good idea of what level of face memory ability you have. Maybe you suspect that you might have prosopagnosia (disability in face memory) or you might just want verification that you are what we call normal. When I finished the test I was given my own score and also an average score, so I guess you could use this test to compare yourself against the norm. I am not sure whether the researchers might object to curious people rather than potential super-recognizer study subjects doing the test. If that is a problem, they can let me know. I am also not 100% sure whether or not all scores from doing the online test are used as anonymous research data, as is the case in some websites that offer online neuro-cognitive testing, or is this testing merely used to screen eligible candidates for an upcoming study of super-recognizers. I suspect the latter. You should contact the researchers yourself if you have any questions.

This is an important new opportunity to undergo a test of face memory ability because this test is a version of the Cambridge Face Memory Test, a test that has a well-deserved reputation as a scientifically valid test of face memory, and it has been quite a long time since I’ve been aware of any version of the CFMT being openly available to take through the internet. This test appears to be a third version of the CFMT, the first being a 72 question short form, the second version being the 102 question long form, and this version being another 72 question version but with new faces (I’m pretty good at judging these things) that are computer-generated. All versions have all male faces. I recall reading somewhere that the faces used for the first version of the CFMT are based on real American Caucasian people. It’s probably a good idea to use computer-generated faces for the latest version, to avoid the possibility that real people might be stopped in the street by super-recognizers exclaiming “Hey! You’re one of those blokes in that face recognition test!”

For people hoping to find a way of documenting their own status as a super-recognizer this is an opportunity to do a scientifically credible test of face recognition and also get access to your own written score in that test along with an average score, but be advised that I did not automatically get any printed statement verifying that my score was in the range of super-recognizers, and I did not notice any printed range of scores for super-recognizers given anywhere in the testing. It was pretty obvious from the results page that my score was in the elite range though. I have been advised by someone who should know that a score of 69 is considered to be in the super-recognizer range. I found this test to be harder than the first version of the CFMT, and I suspect that super-recognizers might find that they don’t bump their heads on a ceiling with this test. Information that I have at hand suggests that the average score on the new 72 question version of the CFMT is lower than the average score on the old 72 question version. For super-recognizers who are eligible and willing to take part in the study after they have done the screening test I guess there might be further opportunities to document their status as a super-recognizer and meet researchers, but I can only guess. At least your participation would give you the right to refer to yourself as a “Citizen Scientist”, which sounds fairly impressive. As I live in Australia I will not be able to participate in the study so I can’t advise you where it all leads. My best tips for people interested in documenting their score are to follow the instructions carefully, be ready to take your own full-colour print-outs of any screen with your score on it and if you have questions contact the people behind the test.

This isn’t the first time that I’ve been asked by a researcher from an overseas university to help with recruiting supers for research studies, and I’m happy to help with genuine requests because I like to see science moving forward and I know that participating in research can be interesting and sometimes rewarding. The website TestMyBrain is associated with many genuine researchers of social psychology, neuro-cognition, visual perception, face perception and various interesting and important things, including researchers from the Social Perception Lab at Dartmouth College and researchers from the prestigious Harvard University. Researchers linked to the website appear to be generally based at universities and colleges in the United States of America.

In case you are curious, my score on the new test is 69 out of 72 and I have been advised that my score is in the super-recognizer range. As I have written about in old posts at this blog, I have also done the earlier versions of the CFMT. The first version of the CFMT was either the first or the second face memory test that I ever did, and I was amazed at the time to get perfect scores on both online tests I did that day. I had gotten 72 out of 72 on the old version of the CFMT, but till then I had no idea that I was a super-recognizer. In 2010 through an Australian university I did some face memory tests in person and I firmly believe one of those tests was the long form of the CFMT, my score in that one given as “96%”, which presumably means I got 98 correct out of a possible 102 correct, which is well within the super-recognizer range, based on data about supers from the 2009 journal paper that launched the concept of the super-recognizer. As I’ve stated earlier, I have at hand data that indicates that the new CFMT is more difficult than the earlier version of the same length. The norm for the first version was given as 80% correct face recognition (presumably an average score of 57.6 out of 72) while the average for the new version is currently cited as a score of 52.49 (out of 72). So it appears that the CFMT has become more difficult while my face memory ability has not measurably changed in the four years since it was first tested.

I wish the researchers planning to study supers the best of luck and I look forward to reading a published report of what they find. We are all working to help people and to advance scientific knowledge, and those are for sure two noble causes.

Cambridge Face Memory Test- how long is it?

The original short standard version had 72 items or questions in it and the long version had 102, but I have recently read that new versions have been created, and I don’t know anything about them. There is also a children’s version.

This is a link to the journal paper that introduced the concept of the super-recognizer, and you can read about the use of both original versions of the CFMT in this paper: http://www.faceblind.org/social_perception/papers/russell09PBR

Randomly finding studies that have un-noted super-recognizers in them

Don’t assume that if face memory researchers find that they have study subjects who get scores typical of super-recognizers that they will note this fact in the paper or will interpret the data to inquire about the characteristics of super-recognizers. In random internet clicking I keep coming across studies that have super-recognizers in them, and which also appear to have no comment in them about this finding.

Joshua M. Davis, Elinor McKone, Hugh Dennett, Kirsty B. O’Connor, Richard O’Kearney and Romina Palermo Individual Differences in the Ability to Recognise Facial Identity Are Associated with Social Anxiety. PLOS One. Published: December 14, 2011. DOI: 10.1371/journal.pone.0028800. http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0028800

There are two subjects in this study who got scores in the super-recognizer range (71-72 in the short form of the Cambridge Face Memory Test) and three who were pretty darned close with scores of 70, and there were also eleven subjects who scored in the prosopagnosia range of 42 or less in the CFMT short form.

The study measured social anxiety and other things, so what were the findings in relation to the supers and social anxiety? From what I’ve read the cut-off point for social phobia is 36 or more on the SIAS, and none of the supers were anywhere near that, but that is also true for the majority of the subjects who got face memory scores in the prosopagnosia range. Plenty of study subjects got SIAS scores in the social phobia range, but only one of them also had a CFMT score in the prosopagnosia range. Social phobia clearly isn’t explained by issues in face recognition. The study did find a weak correlation between poorer face memory and social anxiety, but I’m surprised that a stronger positive relationship was not found because I think face memory must be pretty important in social functioning. I offer the Dunning-Kruger Effect as an explanation for the weakness of the correlation found. People who don’t know often don’t know what they don’t know and maybe even don’t know that they don’t know. Ignorance is bliss, so they say, and I see evidence of it all around me every day.

In my opinion, the alternative explanation offered to explain the slight correlation between social anxiety and poorer face recognition, that social anxiety could cause the development of poorer face recognition ability, seems unlikely. Kids who are scared of other kids would surely want to keep tabs on who is who, because such kids are likely to be bully-magnets, and certainly not all kids are bullies. I would think in such a situation, a child would have great personal interest in telling the difference between bullies, allies and the general mob, and face memory would seem to be the best tool for this task. That’s just a theory, and often sensible-sounding theories are totally counter to reality, so more research is definitely desirable on this question.

Here’s another study that has super-recognizers in it, but which doesn’t specifically mention or discuss super-recognizers or superrecognition:


[this article to be completed later]

New paper about study of face processing in developmental prosopagnosia on oxytocin

The paper is open access, so you don’t need to pay to read the whole thing. Is “face processing” the same thing as “face memory” or “face recognition”? When I’ve got more time I’ll have a good look at this study and see. I have noted that this is a quite small study (10 DPs, 10 controls), so let’s not get too excited about the findings.


Two bits of interesting information about the Cambridge Face Memory Test can be found within this paper.  The authors advise that some people with developmental prosopagnosia can achieve a normal score on the CFMT by using “effective compensatory strategies”. I’m curious about how that is done, because I thought the CFMT was pretty much cheat proof. It is also revealed that two new versions of the CFMT were created for this study.

I plan to write more about this paper but right now my garden requires attention. And after that the turquoise coastline lined with fine white sand near where we live will require attention.