Bias in recruitment and selection has taken a lot of stick in recent years. Organisational standards implores that policies and procedures inject a heavy dose of being mindful when it comes to equality and diversity in recruiting new candidates. Of course this is not just what the law requires, but also on how the organisation wants to reflect, shimmer and shine over others as being accepting, enabling and inclusive. ‘Bye-bye’ stuffy corporates where one ticks the box in recruiting a mild mix of minority, gender, disability and age variants for statistics and ‘hello’, we employ based on experience, competency levels and cultural fit. I mean who judges on looks alone when we are all so very PC minded?
So why are we walking on eggshells when it comes to exposing the real difficulties in reducing bias? – come on we are HUMAN – full of stereotypical behaviours and hidden perception’s and we shouldn’t be ashamed of this, it’s how we are programmed – even the most trained and experienced recruiters and assessors are still at risk of these latent traits seeping through (Dipboye, Franklin and Wilback, 1975).
OK – so we can’t change being human and our needs for a strong work-force, better ethics and best practice methods for recruitment & selection are still very much a necessity (well not until our new found – hopefully friendly – co-bots will take over all jobs and eventually the world), so while we still have a stab at making this work we may as well try our hardest – the human way.
Goldberg (2006) claimed that we are all attributed with ‘perceived similarity’ and ‘interpersonal attractiveness’ tendencies, this in layman’s (or women’s) terms means that assessors and recruiters cannot completely diminish or control how predictor ratings in assessment can eliminate subjective bias in sex, age, minority or disability. How we select is construed with intentional and unintentional discrimination. We have seen this time and time again, posts on ageism (older and more experienced is less likely to be offered the job vs. younger whom are less likely to be promoted (James et al., 2013)), gender selection (may the best man win!), minority (different cultural backgrounds may not fit with overall perceived organisational culture (Hofhuis et al., 2013)), disability and disfigurement, for example, facially disfigured applicants are less likely to be selected, even if they are more equip for the role (Stevenage and McKay, 1999).
The Harvard Implicit Association Test (IAT) (https://implicit.harvard.edu/implicit/takeatest.html) is an example of how you can test your own preconceived attitudes which are then associated within a range of different concepts or implicit/explicit stereotypes. However the study itself highlights that the IAT should be used purely for educational/research purposes and not to be misused, for example it would be totally unethical to use this test to judge a potential candidates own biases!
The point is, although we talk the talk about reducing bias, we are only marginally stepping towards re-educating how we assess, the techniques used and implementing real valid and sustainable change in the process.
Traditional assessment and interviewing strategies are moving away from the legacy paper-based and face-to-face engagements and are being replaced by technological advancements, such as online tests & questionnaires, game-based assessments, stronger situational judgement tests (SJTs) and more streamlined interview processes, for example, online and video interviewing techniques. All are innovating the landscape by ways to help reduce subjectivity in the recruitment process.
So, you may ask, how can e-assessment environments further improve this quest? Well I am glad you asked! Assessment centres are arguably one of the stronger ways to assess candidates, they can measure more situational outcomes between and within exercises and participants (Burnett, 2015), thus helping identify capable participants much more quickly (as you will have a group results rather than having to sift through individual assessment outcomes or reports) However nothing is perfect, sadly assessment centres have their imperfections just like the rest of us. Firstly it’s the cost, on average each candidate is estimated at £311, but some employer’s reveal costs can be as high as £1,500 – £2,000 (Xpert HR, 2011). Secondly, it’s the length of time it takes to manage resources, administrate and host the assessment centre, adding on the expenditure.
E-Assessment environments such as t-PHIs & VirBELAs wonderful platform J can help alleviate both the resource time and cost – good news right? Well there is more….
By placing candidates in a virtual environment – you could potentially reduce bias – Candidates have the option of creating their own avatar, whereas ethnicity, gender, age and disability can become neutral. Exercises are set up in-world to test competency AND behavioural levels, for example, imagine a candidate passed with flying colours the performance management assessment but failed on how they ‘virtually’ engaged with other team players in the same exercise? Yes guys, richer data awaits…
Attributes of an Avatar
Although we may seemingly be on course to becoming more aware of attractiveness/subjective bias with the introduction of tools such as t-PHIs and VirBELAs e-assessment environment, it is possible that avatars are still at risk of being unconsciously judged. The digital world has enabled us to create and recreate ourselves online. However those that have had the experience of platforms such as ‘second life’ (http://secondlife.com/) or ‘weeworld’ (http://www.weeworld.com/) would understand that we are more inclined to create avatars that represent our true selves. Be that the colour of our skin, hair or eyes, gender and approximate age range. With this in mind it wouldn’t be unconventional that how an assessor responds to an avatar will be similar to how they form first impressions if they met face-to-face, e.g., based on certain attributes such as physical appearance (Fong and Mar, 2015) and/or an avatars perceived agreeableness – for example do they come across fair, fun, friendly or distant and unengaged. It could also be argued that the lack of visual cues (such as a genuine smile) could reduce the human authenticity of the avatar candidate.
On the contrary those limitations can also be the strength of what e-assessment environments and avatar-based candidates can bring. Attractiveness bias can potentially be reduced as face value is replaced with all candidates being assessed for how they should be assessed – for being the best person for the job!
I recently asked HR & recruitment experts their views on how e-assessment environments can help reduce managerial bias and the feedback was fair. Overall virtual assessment environments can create wider audiences, for example can be used to host L&D engagements, leadership development as well as assessment, costs can become more realistic and proportionated as well as reduce resource time (thumbs up) but there is a consensus that face-to-face interviews are still very much an activity that should belong within the recruitment process to assess the person and their potential fit.
All very much valued thoughts. But with research still pointing towards the existence of subjective (or unconscious) bias in recruitment and selection we are faced with the challenge – do we train up assessors to reduce their own negative stereotypic reactions? And if so how? Recruitment techniques are very rarely frequently implemented, in an ideal circumstance, randomised methods and techniques should be utilised, however the truth is even simple randomisation techniques lack consistency (Putka and Hoffman, 2012). A suggestion would be to train two assessors per one candidate over 36 hours to reduce observer bias, although findings suggest a reliable variance between inter-raters, the realistic implications is that employers do not have the time to train and the cost of training makes this suggestion less commercially viable to resolve the bias issue (Putka and Hoffman, 2012).
The quest for reducing managerial bias is still very much apparent, more research needs to be done to better understand the effects subjective bias has in different concepts of the recruitment journey and the advancement of different technological developments allows us to start building upon that knowledge. However one thing is clear, candidates are getting tired of responding to static tests and questionnaires – and folk are getting better at doing this. They want to have transparent and fair experiences, be ‘wowed’ by what employers are offering, whilst employers need richer data to make better and valid decisions at a faster rate without having to think ‘have we been balanced both ethically and diversely in our processes’?
Our very own e-assessment environment begins to challenge that quest. By allowing us all to engage with human behaviour within different contexts we will begin to gain a richer understanding of our own capabilities, judgements, weaknesses and strengths. So before co-bots laser-beam our desks away (they may/may not have laser-beams for eyes) and come up with a better solution to the bias quest, we should continue learning with the means we know best, through trial and error, which is of course the HUMAN way.
On that note, what you waiting for? Now is the time to challenge the bias issue by bringing innovation into your world!
Contact Us today for a virtual tour – prepare to be amazed!
t-PHI on Immersive Assessment
At t-PHI we believe that too much time is focussed on candidates responding to static statements about how they feel they should answer a question and not enough focus on how they actually behave within a situation. Immersive assessments such as the e-assessment environment, starts to define these differences and measures behavioural outputs by using psychological science, psychometric and assessment techniques with innovative technological solutions.
Burnett, G. M. (2015). Situations and their influence on the measurement of latent traits. [Online]. Available at: http://epubs.surrey.ac.uk/807155/ (Accessed 2015)
Dipboye, R. L., Fromkin, H. L., & Wiback, K. (1975). Relative importance of applicant sex, attractiveness, and scholastic standing in evaluation of job applicant resumes. Journal of Applied Psychology, 60, 39-43
Fong, K., Mar, R.A., 2015. What Does My Avatar Say About Me? Inferring Personality from Avatars. [Online]. Available at: http://psp.sagepub.com.libezproxy.open.ac.uk/content/41/2/237.full.pdf+html (Accessed June 2016)
Goldberg, C.B. (2006) Relational demography and similarity-attraction in interview assessments and subsequent offer decisions: are we missing something? [Online]. Available at: http://www.emeraldinsight.com/doi/abs/10.1108/hrmid.2006.04414cad.005?journalCode=hrmid (Accessed June 2016)
Hofhuis, J., van der Zee, K. I., Otten, S. (2013) Measuring employee perception on the effects of cultural diversity at work: development of the benefits and threats of diversity scale. [Online]. Available at: https://www.researchgate.net/publication/271659470_Measuring_employee_perception_on_the_effects_of_cultural_diversity_at_work_development_of_the_Benefits_and_Threats_of_Diversity_Scale (Accessed June 2016)
James, J.B., McKechnie, S., Swanberg, J., Besen, E. (2013) Exploring the workplace of intentional/unintentional age discrimination. [Online] Available at: http://www.emeraldinsight.com/doi/abs/10.1108/JMP-06-2013-0179?journalCode=jmp (Accessed June 2016)
Putka, D.J., Hoffman, B.J., (2012). Clarifying the contribution of assesse-, dimension-, exercise-, and assessor-related effect to reliable and unreliable variance in assessment center ratings. Journal of Applied Psychology, 2013, Vol. 98, No.1, 114-133
Stevenage, S. V. & McKay, Y. (1999). Model applicants: The effect of facial appearance on recruitment decisions. British Journal of Psychology, 90, 221-234
Suff, R., 2014. Assessment Centres: well worth the money. [Online]. Available at: http://www.xperthr.co.uk/blogs/employment-intelligence/2011/09/assessment-centres-well-worth/ (Accessed June 2016)