School of Human Sciences
Face and Voice Recognition Lab
Institute of Lifecourse Development
University of Greenwich
London
www.superrecognisers.com
super-recognisers@Greenwich.ac.uk
Twitter: @GRecognisers
19 August 2021
Greenwich Face Identity Tests: Pilot Results
Katie Read (Student Research Assistant)
Josh Davis (Professor of Applied Psychology)
Thank you to all participants who recently participated in the Greenwich Face Matching 100 Test (GreMatch100) and Greenwich Face Memory 50 Test pilots (GreMem50) (note: we had already conducted five pilot tests of the Greenwich Face Matching Test, and two pilot tests of the Greenwich Face Memory Test – these involved over 2,000 participants).
We always receive large numbers of e-mails requesting information about test difficulty and how their scores compare to others. As we could not provide this immediately in feedback at the time (we did not know), we have done so here knowing that it encourages participants to contribute to additional research.
Below is a brief but incomplete description of the tests and summary of scores collected so far. We intend to publish the full data at a later date, so cannot provide too many details now (journals do not normally accept submission if data are previously published). However, the main reason we asked participants to take these pilot tests is that we wish to use them as a measure of ability within different contexts in our future research.
Greenwich Face Matching Test (GreMatch100)
This 100-trial test measures the ability to distinguish the identity of pairs of photos of faces from multiple ethnicities. Participants must decide if the images are of the same or of a different face. Participants provide confidence in each decision. The images were obtained between 2015-2020 and were either taken at the University of Greenwich or were uploaded by participants. The images were designed to allow for natural variability commonly found in photos (from selfie to posed portrait).
Greenwich Face Memory Test (GreMem50)
In the first phase of this test, participants view a series of four photos of the same person in the first phase. All were taken in different contexts. They are asked to learn the faces. Unlike most of the other tests of memory in our battery, participants have as long as they like to study the faces. After a break, which involves a complex distraction task involving facial photos, in Phase 2, participants view 50 facial images sequentially and have to decide which are ‘old’ and ‘new’. Proportions of old and new trials are not equal in order to match likely real-world experiences.
Participants
We sent invites to a proportion of our 47,000+ volunteer participant database. Some invites were chosen at random from the entire database, while many were random selections of known super-recognisers (based on other tests).
The mean scores of those who contributed (by 1 August 2021) on the Glasgow Face Matching Test (n = 852, M = 37.52 out of 40, SD = 2.41), Kent Face Matching Test (n = 824, M = 31.31 out of 40, SD = 3.326), Cambridge Face Memory Test: Extended (n = 852, M = 86.96 out of 102, SD = 10.59) are higher than is typically found in research suggesting that these participants, as a whole, possess better-than-typical ability. (note – not all database participants have taken the Kent Test yet).
Therefore, those who contributed to this project may have scored higher than would be typical on the two Greenwich tests as well. As such, the scores are unlikely to be representative of the general population. We will comment on this in the journal article. We collected additional data from participants at the same time (e.g. confidence) that we will report in full.
Results
Please find in Figure 1 a histogram of scores on the Greenwich Face Matching Test (out of 100), and in Figure 2 a histogram of scores on the Greenwich Face Memory Test (out of 50) – these data are not yet cleaned for duplicate entries, unmatched participant codes, or other anomalies. Nevertheless, you should be able to compare your score with those reported here to give you a rough idea of where you stood.
Figure 1: Greenwich Face Matching Test (GreMatch100) scores (n = 886 – some duplicates included)
Figure 2: Greenwich Face Memory Test (GreMem50) scores (n = 873 – some duplicates included)
