Although it won’t work for every employer, automated personality testing could revolutionise how we handle volume recruitment, says Dr Paul Barrett
Social networks have clearly already had a huge impact on the HR function. There are the obvious things, such as how LinkedIn and Facebook have made it easier to directly identify interesting candidates, and enabled more organisations to cut the cost of their recruitment processes. Then there is the growing use of private social networks, such as Workplace by Facebook, for employee engagement and communication.
These tools have already proved their worth as part of the talent management process and now a new wave of automated candidate assessment and profiling is emerging – again supported by social media.
The use of social media data for candidate assessment came to the fore recently following claims made by Michal Kosinski, one of the leaders in personality assessment using social network activity linguistic analysis, to verify the effectiveness of his methodology. His technology effectively scrapes data from an individual’s ‘digital footprint’ and uses this to build a profile of their personality based on the OCEAN framework for recruitment purposes.
According to Kosinski’s research team at Cambridge and Stanford Universities: “With a mere 10 ‘likes’ as an input, the model could appraise a person’s character better than an average co-worker. With 70, it could ‘know’ a subject better than a friend; with 150 likes, better than their parents. With 300 likes, Kosinski’s technology could predict a subject’s behaviour better than their partner. With even more likes it could exceed what a person thinks they know about themselves.”
For HR practitioners, this creates an opportunity to not only automate the identification and targeting of potential talent for their organisations, but to extend this to automating the initial vetting process too, using profiling data derived from social media. Although HR already uses various applicant-tracking systems, a significant amount of human intervention is still required, whereas with social media profiling, finding candidates for a job could be done by an autonomous, end-to-end system.
This has some clear benefits for HR practitioners, especially those working in large organisations and receiving vast numbers of job applications. For instance, it could completely remove the need to engage with personality questionnaire assessments. Research has shown that automated assessment systems are able to predict personality traits just as accurately as the standard self-report personality questionnaires. Candidates simply have to provide sufficient information so that their digital footprint can be uniquely identified.
Once identified, the tool would ‘score’ the personality attributes for a candidate and run these against one or more job-role target profiles that have been created by HR or even by ‘big data analytics’. Depending on the outcome, the tool would then either report back to HR if a candidate’s scores passed the screening process, or send a polite message informing the candidate they have not been successful.
It’s potentially a completely seamless and totally automated ‘pre-screening process’. All the candidates can be handled autonomously and HR only needs to be notified of candidates who meet predefined bio data and personality filters, leaving more time to focus on applicants who meet the profiled requirements for a job role. Airlines and the hospitality industry, for example, receive thousands of job applications each month and need to find a way to reduce that number to a more manageable quantity and ease the financial and logistical burden.
These developments are very useful for sorting purposes, but social media profiling is unlikely to be much use beyond this recruitment-oriented task because it can only extract information from what is essentially linguistic or a very limited kind of behavioural activity. Plus, it could potentially be exclusionary, as some candidates will have made a conscious decision not to be prolific users of social media. Will they become excluded from segments of the job market as a result?
In addition, when hiring for highly autonomous or leadership roles, detailed assessments will continue to be important as identifiers of a candidate’s psychology. Data scraped from social media could never be a substitute. Longer term we can expect to see the candidate test market split between automated tools used at the top of the recruitment funnel, and very specialised assessments used at the sharp end to really pinpoint cognitive capabilities and potential.
Dr Paul Barrett is chief research scientist at psychological assessment company Cognadev