Introduction to Affinity 3.0 

David V. Glasgow

When in 1995 I combined two computer-based assessments of sexual interest to become Affinity 1.0, I anticipated that this assessment would probably have a life-span of perhaps two to five years.  I imagined that new assessments would quite rapidly arrive offering similar and probably improved functionality.  It is true that a number of new assessment procedures have been developed, but all have approached the task in a fundamentally different way to Affinity.

By the time that Affinity was in development, there was already a very well established literature indicating that there was a significant, if moderate, correlation between sexual interest in the subject of an image, and the time an individual spent viewing the image.  Further, if one imposed a concurrent task on the person viewing an image, their performance might be enhanced or impaired, depending upon the degree of sexual interest in the image being presented.

Affinity was originally designed to enhance and validate the self-reported sexual interest of adults with a learning disability.  I did not wish to use sexually explicit images, and therefore anticipated that the correlation between viewing time (VT) might well be more modest than that reported in the literature (which tended to use sexually explicit images).  Indeed, I thought that it was possible that the correlation would be very small, or absent.

In contrast, it was very evident that the correlation between VT and sexual interest would have to be very high indeed to offer sufficient sensitivity and specificity to use it as a primary measure of sexual interest.  I therefore took the decision to use VT not as a primary measure of sexual interest, but as a concurrent validity measure.  Careful structuring and standardisation of self-report meant that even a very modest correlation could indicate dishonesty if an individual being assessed ranked a simple archetype of a male or female (a line drawing) in a lower position than was warranted by actual sexual interest.

The logic of the Affinity assessment involves five processes:

  1. Eliciting a rank order of archetype drawings which represent the “baseline” of self-reported sexual interest (by sex and 4 age groups).
  2. Gathering ratings of photographic images corresponding to the age and sex categories of the archethyes (10 exemplars x 2 sexes x 4 age categories = 80 images).
  3. Recording the time taken to make the rating of each image.
  4. Examining the data for all three of the above concurrently, both in raw (bar chart) and ipsative (line chart) form.
  5. Interviewing/debriefing the subject of the assessment regarding any inconsistencies in the data, either with respect to individual images, or categories of image.

This is a rather conservative approach to the use of VT, combining it within a multi-measure, ipsative approach to assessment (Glasgow, 2009).  The emphasis is on engagement, enabling systematic self-report, promoting honesty, investigarting and exploring inconsistency.  The resulting data enables professionals to generate formulations relating to sexual interest, including the possible reliability or unreliability of self-report.  It can certainly inform actuarial assessment, but it certainly does not serve as a “paedophile detector,” categorising individuals according to sexual interest determined solely by VT.

An extensive literature on cognitive assessments of sexual interest has emerged over the past 16 years.  This has served to confirm that the conservative approach to using VT within Affinity was entirely justified.  Although research results vary somewhat dependent upon methodology and stimuli used, the overwhelming evidence is that the sensitivity and specificity of VT measures remain insufficient to justify the use of a simple VT test as a means of categorising individuals according to their putative sexual interest.

There have been some extremely promising results from a number of studies, for example using a choice reaction time measure combined with the Not Real People stimuli (Mokros et al., 2011) developed by Pacific Behavioural Assessment (  Mokros et al. (2013) have also evaluated the reliability and validity of the Affinity program.  Banse, Schmidt and Clarbour (2010) have reported considerable progress using VT as part of a “stable” of direct and indirect measures of sexual interest.

Affinity 3.0

Version 3.0 of Affinity is a revision that has incorporated suggestions from many collegial sources.  Perhaps the most obvious change is increased size and quality of images.  The small photographic images of Affinity 2.5, in only 256 colours, were a legacy of Windows 3.1.  The operating system also used an approach to file management which has been entirely superseded in Windows 7 and 8, and Mac OS X.  Affinity now saves results in open text format within the documents folder of individual users.

Another significant addition is the facility to undertake an Affinity-style rating on any set of images stored within a folder on the computer.  This is termed “Open Edition.”  This is in response to requests from users who wanted to obtain detailed self-report and VT data on diverse images, such as various fetishistic objects, violence, or images related to arson.  No claims can be made here with respect to what might emerge from investigations of this sort, but the software now affords to opportunity to easily conduct exploratory studies according to hypotheses generated by users.

The debrief process has been promoted to a more obvious position, permanently prompted by an image representing the currectly selected stimulus in the bar chart view.  The is to encourage the use of the debrief to engage with individuals being assessed, clarify the determinants of the observed data, and perhaps present a final opportunity to acknowledge denied deviant sexual interests.

Whether Version 3.0 of Affinity is the last depends very much on the feedback of users, as well as whether the multi-measure, formulatoroy approach to risk assessment and management prevails over the rather more simplistic single measure, categorical approach.

I remain grateful to all for the enthusiasm, support, and even critical comments of past users and researchers.  In particular, I would like to acknowledge the comments, constructive criticisms and suggestions of, specifically, Mervyn Davies, Gary Reser, Richard Packard, and my business associates, Richard Laws and Carmen Gress.  They have helped shape the revisions of Affinity over the past 16 years.  If users of this version of Affinity 3.0 find it helpful in their work, it is in no small part due to the efforts of these and many other users.


Banse, R., Schmidt, A.F., & Clabour, J. (2010).  Indirect measures of sexual interest in child sex offenders: A multimethod approach.  Criminal Justice and Behavior, 37, 319-335.

Glasgow, D.V.  (2009).  Affinity: The development of a self-report assessment of paedophile sexual interest incorporating a viewing time validity measure.  In D. Thornton & D.R. Laws (Eds.), Cognitive approaches to the assessment of sexual interest in sexual offenders (pp. 59-84), Chichester,UK: Wiley-Blackwell.

Mokros, A., Butz, M., Dombert, B., Santtila, P., Bauml, K.-H., & Osterheider, M. (2011).  Judgement of age and attractiveness in a paired comparison task: Testing a picture set developed for diagnosing paedophila.  Legal and Criminological Psychology, 16, 323-334.

Mokros, A., Gebhard, M., Heinz, V., Marschall, R., Nitschke, J., Glasgow, D.V.,Gress, C.L. Z., & Laws, D.R. (2013).  Computerized assessment of pedophilic sexual interest through self-report and viewing time: Reliability, validity and classification accuracy of the Affinity program.  Sexual Abuse: A Journal of Research and Treatment, 25, 230-258.


Sub Navigation

Contact Us

D. Richard Laws, Ph.D.

David V. Glasgow, M. Clin. Psych.
Head, Product Development

Mail Address

PO Box 23032
Cook Street R.P.O.
Victoria, BC V8V 4Z8

Courier Address

101-230 Cook Street
ATTN: Post Office
Victoria, BC V8V 3X3

Fax: 250.388.5787