15th May 2019, Manchester, Kiki Leutner – Teaching Fellow in Psychology, UCL and Consultant at HireVue.
Games and assessments are rapidly becoming synonymous with each other and games are becoming part of the “Natural Environment” for job applicants.
Games are proving to be a very effective means of evaluating various personality characteristics, in particular when they are a form of guided psychometric assessments.
When choosing technology or AI driven assessments, it is key to distinguish between those that use a trait approaches vs those that use a performance approach to finding the most suitable applicant. Trait approaches typically use new technology such as games to predict scores on validated psychometric tools, and are optimised for convergent validity.
Evaluating and predicting performance can be more challenging, but shows great results if used in the right settings. The key factor in performance based approaches is the availability of robust performance data from several hundreds of employees in the same or similar roles. If this data is available, any form of performance based assessment should start with whether an organisation is happy with its existing workforce. If so, performance data for existing roles can be extremely valuable and used as benchmarks if the data is sufficiently rigorous. If not the data can still be useful but a thorough analysis of desired changes needs to be undertaken. Otherwise, there is a danger that all an organisation will be achieving is a “cloning” effect of its current workforce, missing out on the benefits of a diverse workforce.
Any psychometric assessment can produce undesirable Adverse Impact. This is particularly true in the case of cognitive ability tests. In addition, candidates can suffer a negative experience with such tests, as they produce anxiety and put people on the spot. Games provide a valid alternative to traditional measures of cognitive ability. They have several advantages, especially a better user experience that is less anxiety inducing. Assessments are also accurate and the underlying machine learning models enable them to be optimised to reduce adverse impact. Additionally
- they benefit from shorter assessment times
- Access is improved through using colour blind versions, or adjustments for different needs such as dyslexia.
- The user experience is improved, which results in higher completion rates of up to 95% in life projects.
Identifying the best provider is challenging as the market is moving so quickly, but examples of issues organisations need to consider when selecting a provider are:
- Does the assessment match employer branding?
- Application process, timing and degree of rigour
- Is the assessment perceived to be ethical in the locations it will be used in?
- Is the assessment fair, e.g. are large samples available to demonstrate adverse impacts on the process?
- Does the process correlate with existing measures of personality/IQ or the relevant traits it claims to measure?
- Does the assessment correlate with job performance in client projects?
- Are highly qualified (e.g. PhD level) psychologists involved in game development and are these people available for discussion if necessary?
- Is there any peer review of the process and is there a record of user experience?
In summary, the Games Based Assessment process has proved that it can be engaging and immersive, producing hugely improved information on situational judgment, measurement of emotion and soft skills. Building a relevant game can then be a valid process for the assessment of cognitive ability at a level which has not before been achievable. The rapid increase in sophistication combined with overwhelmingly positive feedback from users means that Games-based Assessments have firmly established themselves as an accepted and reliable part of the assessment and selection process.
Please refer to the presentation given by DR Kiki Leutner here.
Report by Richard Taylor.