Human Assets: Helping you choose, develop and engage winning talent.

For You


Accuracy of candidates’ self assessment and assessment centre decisions

A study published in December 2000 by researchers from Nottingham University explores the link between the accuracy of self assessments and the likelihood candidates will be accepted in graduate assessment centres. Workplace behaviours thought to be associated with the ability to produce accurate self assessments are improved team behaviours, adaptability, greater willingness to seek and use feedback and stronger working relationships. Previous research has suggested that those who are accurate in self assessments of performance are more effective in the job and the assessment centre.

During the assessment centres in this study candidates rated their own performance on a nine point scale as used by the assessors. The study found:

  • On three exercises of the six exercises at the centre, the successful candidates’ ratings were generally found to be closer to the ratings produced by objective measures (assessor ratings, psychometrics) than unsuccessful candidates. The three exercises were two group exercises and the structured interview.
  • No difference was found between accepted and rejected candidates on the written ‘in-tray’ exercise.
  • In many cases, but not all, the high performing accurate raters were those who were accepted.
  • The researchers also looked at absolute accuracy rather than just the degree of inaccuracy. In group exercise two 1 (a discussion), but not group 2 (a discussion and presentation) or the interview, absolute accuracy could be used to discriminate between accepted and rejected candidates – successful candidates’ assessments were likely to match ratings by other objective measures.

Several explanations can be given as to why self assessments can assist predictions:

  • Better candidates can assess their own performance because they have more self awareness. Self awareness helps them to demonstrate their strengths and manage their behaviour.
  • Good candidates recognise the criteria on which they are being evaluated or what is required to succeed in the organisation and thus are able to judge their own performance based on a framework which is similar to the assessors.
  • Better performers do not necessarily produce ratings in agreement with assessors across all exercises.

The nature of the exercise matters:

  1. In a group discussion candidates need to receive, judge and act on feedback not just demonstrate competence. Therefore candidates who can assess their own behaviour can monitor their actions and effectively regulate behaviour. Exercises which provide continuous feedback may therefore enhance self assessment and discriminate between high and low performers.
  2. Observers also place more emphasis on group performance and observable data when making assessment centre decisions (Anderson et al 94). Therefore the group context allows those who are good at self assessment to have an impact on the results which contribute most to assessment centre decisions.
  3. The difference between group exercise 1 and group 2 is important. Group 2 involved a short presentation by each candidate, whereas group 1 was purely a discussion. The presentation in group 2 does not require self monitoring and regulation to the same degree to do well. Therefore those who are perceptive in their self assessments in interactive situations have more opportunity to inform and develop their self assessment, and thus get a closer match to assessor ratings (absolute accuracy), in group 1 than group 2.
  4. The behaviours required in interactive exercises could be more amenable to self assessment than the cognitive processes required in individual problem solving tasks e.g. in-tray exercises. This study found no difference in the self assessments of high and low performers in relation to assessor ratings for the in-tray. The ‘intelligence’ required to perform well in a written exercises does not necessarily equate with accuracy of self assessments.
  • The study found that some low performers who achieved accurate assessments were also accepted by the centre. It would be interesting to explore how they used their self assessments to ultimately be successful. For example if they perceived their performance to be poor in the first group exercise did they make adjustments for the second exercise? If they felt they had not demonstrated their strengths in the group exercises did they try harder in the structured interview?

Impact on selection

The fact that successful candidates were more accurate than unsuccessful candidates in three of the exercises does provide interesting food for thought. Should we be using self assessments to a higher degree in assessment centre decision making? At this stage it is probably sensible to introduce a self assessment element and to monitor the results of these assessments in line with recruitment decisions.

It is important to at least be aware of the importance of management self awareness in performance on the job. Try to ensure your centre contains exercises which allow candidates to draw on self assessment abilities to monitor their behaviour accordingly – and thus demonstrate positive behaviours to the assessors.

The concept of emotional intelligence and its contribution to success as a manager, seems to link to this research in terms of self regulation and self monitoring. However, be wary of simply introducing the recent EI tests on the market to gauge candidates ‘self awareness’. These tests might have flaws of their own.

The advantage of this study is that it also looked at self assessment compared with other ‘objective’ measures such as psychometrics. Accurate self assessments were found to be a better predictor of whether a candidate would be selected than psychometric tests.

At the present time it may be more appropriate to ensure self assessment plays a larger role in development, particularly to assist skill development.


Randall, Ferguson & Patterson (2000) Self Assessment Accuracy and Assessment Centre Decisions. Journal of Occupational and Organisational Psychology, Vol 73 (Part 4).


August 2013

Copyright © 2013 Human Assets Ltd. Registered office: 28 Ulysses Road, London, NW6 1EE. Registered in England. Company number 2075093