When the job hunts you

Linda Kaye on putting our online data to work.

It's 2040. You're still working in Psychology, but it has changed. How?

Twenty years ago, in 2020, we were starting to understand that the information we were posting online may have consequences. In fact, many of us felt we were revealing too much about ourselves, and as a result putting ourselves and others at risk. Although we recognised that different types of information were better suited to be communicated either publicly or privately in these online spaces, what we didn’t fully appreciate was that any information posted online was privy to a third-party audience. ‘Big tech’ was always watching, using our online data to learn more about us. It came to know as much about us as our friends, family and colleagues. It mined our internet search histories, ‘Likes’ and ‘Shares’ of content from social networking sites, friendship connections, membership of group pages, and commercial purchasing behaviours.

For psychologists, however, big tech’s data gathering mission bore fruit in the advancement of psychometric assessment testing in the recruitment process. This has become widely-known as ‘auto-generated psychometric profiling’ (AGPP). We are no longer reliant on time- and resource-heavy processes of organising recruitment assessment days and having to administer and analyse psychometric tests (which we recognised were often subject to social desirability). Instead, we can understand a potential candidate’s personality via their data from ‘big tech’, which allows us a much more efficient and accurate means of garnering their psychological profile to support the ‘matching and suitability’ process for a given job opportunity.

How it works
AGPP supports both potential job candidates and employers, and has been fruitful for the capacity-building of recruitment agencies. A potential job candidate or job seeker signs up to an AGPP-compliant recruitment agency. The agency has access to their psychometric profile which is updated in real-time, garnered from their online data. When new jobs are advertised anywhere online, the agency automatically corroborates this information, and the potential candidate is notified of a ‘match-fit’ to a given opportunity.

Back in 2020 this matching might have happened on the basis of skills and experience, but this delves deeper. The match-fit is based on an automatically-generated similarity fit score between the candidate and the desired psychometric profile of the role. All employers are required to provide the recruitment agency with a psychometric competencies framework for any given job advert.

A potential candidate can then choose to accept or decline the match. If accepted, the employer is then sent a summary of the candidate’s psychometric profile and similarity fit score. They can then choose to proceed or reject any subsequent action, which may or may not include inviting a candidate to submit a formal application.

This has supported the recruitment process in a number of ways to assist both employers and candidates. It has reduced the need for time and resource-intensive assessment days for psychometric testing. The AGPP process is generated prior to a candidate even applying for a job. Only the most psychometrically-suitable people will be personally invited to apply, and any rejected candidates will not have to waste anyone’s time, including their own, on formally applying.

Contributions from cyberpsychology
The field of cyberpsychology has pioneered the evidence base underpinning this advancement. We know that our online behaviour reveals insights into our personalities, and that use of language online impacts judgements of others’ personalities. Specifically, language behaviours such as the use of text-speak, vocabulary usage, writing errors, spelling, and pronoun usage have been found to affect impressions of people’s personalities (Boland & Queen, 2016; Darbyshire et al., 2016; Ellison et al., 2006; Fullwood et al., 2015; Kaye et al., in press; Queen & Boland, 2015; Vignovic & Thompson, 2010; Wall et al., 2016). Even more so, symbols such as emoji, which can really only be gathered from naturalistic online behaviours such as on online social networking sites, relate to impressions of a target’s personality, and importantly, in respect of some personality traits such as openness, these are indeed accurate judgements (Wall et al., 2016). Other online behaviours which also relate to personalities are photo-related content and behaviours (Darbyshire et al., 2016; Eftekhar et al., 2014; Hancock & Toma, 2009). For example, personality judgements are based on how users self-generate online albums and the amount of photos they upload (Eftekhar et al., 2014). All this has helped us understand that aspects of human personality are evident in our online behaviours, and importantly that these enable us to make accurate impressions of others.

We have come a long way since the old days of employers using their own initiative to do random screenings of candidates’ social networking information as an indication of their suitability for jobs. Not only was this non-standardised, it was also largely subject to human biases. Our auto-generated techniques eradicate these issues, and provide an objective, bias-free, standardised alternative. Verification metrics are an important part of that, whereby behaviours are corresponded and verified to ensure they are accurate of an individual. This includes personal identification criteria (e.g. fingerprint, facial recognition) to ensure that these are a true record of that individual and not someone fraudulently posing as them; and social verification via peer endorsement ratings on platforms such as social networking sites, to ensure that content is deemed representative of them. These recognition criteria and ratings provide a confidence metric to corroborate an individual’s auto-generated psychometric profile.

Reflecting back
The role of psychologists in these advancements has been critical. We were vocal on the ethical assurances needed to align with personal privacy while serving a purpose for the public good. Two main motivations convinced us of that. First was the termination of physical Job Centre provision based on the political agenda that ‘unemployment’ was not a discrete identifiable description, but instead a transitory experience which was better supported proactively through recruitment agencies. Occupational psychologists therefore highlighted the need for better systems to support those out of work to find suitable job opportunities, hence the introduction of AGPP. Secondly, the ‘open data’ movement helped us recognise that ‘data’ included that of ourselves, and fuller transparency was required for the purposes of supporting the public good. This required us to reconceptualise what we understood as ‘private’; instead, big tech classified our online data as ‘unrestricted’ or ‘restricted’, in collaboration with the security and defence sector.

However, reservations remain. Even in 2040, we live in a world which includes poverty and there is still a minority of individuals who do not have access to the latest smartphone technology or regular internet connectivity. These are people who remain largely private yet restricted in the traditional meaning of the term, by the fact that AGPP is not yet fully inclusive.

Despite this, AGPP has remained widely popular and successful in the recruitment and employment sector. Employment retention has increased threefold in the last five years, highlighting that this use of our data may be one of a number of factors which helps people find jobs which are more suitable to their psychological profile, therefore increasing their job (and, indeed, life) satisfaction.

- Dr Linda K. Kaye is a Senior Lecturer in the Department of Psychology at Edge Hill University. She is currently Chair of the British Psychological Society’s Cyberpsychology Section. [email protected]

Illustration by Nick Taylor


Boland, J.E. & Queen, R. (2016). If you’re house is still available, send me an email: Personality influences reactions to written errors in email messages. PloS One, 11(3), e0149885.

Darbyshire, D., Kirk, C., Wall, H.J. & Kaye, L.K. (2016). Don't Judge a (Face) Book by its Cover: Exploring judgement accuracy of others' personality on Facebook. Computers in Human Behavior, 58, 380-387.

Eftekhar, A., Fullwood, C. & Morris, N. (2014). Capturing personality from Facebook photo and photo-related activities: How much exposure do you need? Computers in Human Behavior, 37, 162-170.

Ellison, N., Heino, R. & Gibbs, J. (2006). Managing impressions online: Self-presentation processes in the online dating environment. Journal of computer-mediated communication, 11(2), 415-441.

Fullwood, C., Quinn, S., Chen-Wilson, J., Chadwick, D. & Reynolds, K. (2015). Put on a smiley face: Textspeak and personality perceptions. Cyberpsychology, Behavior, and Social Networking, 18(3), 147-151.

Hancock, J.T. & Toma, C.L. (2009). Putting your best face forward: The accuracy of online dating photographs. Journal of Communication, 59, 367-386.

Kaye, L.K., Wall, H.J. & Hird, A.T. (in press). Less is more when rating Extraversion: Behavioural cues and interpersonal perceptions on the platform of Facebook. Psychology of Popular Media Culture.

Queen, R. & Boland, J.E. (2015). I think your going to like me: Exploring the role of errors in email messages on assessments of potential housemates. Linguistics Vanguard: A Multimodal Journal for the Language Sciences, 1 (1), 283-293.

Vignovic, J.A. & Thompson, L.F. (2010). Computer-mediated cross-cultural collaboration: Attributing communication errors to the person versus the situation. Journal of Applied Psychology, 95 (2), 265-276..

Wall, H.J., Kaye, L.K. & Malone, S.A. (2016). An exploration of psychological factors on emoticon usage and implications for judgement accuracy. Computers in Human Behavior, 62, 70-78.

BPS Members can discuss this article

Already a member? Or Create an account

Not a member? Find out about becoming a member or subscriber