SJTs vs Assessment Centres: Which Is Better at What?

Studies You Should Know: SJTs vs Assessment Centres: Which Is Better at What?

The Question

Two of the most widely used tools in sophisticated selection systems are situational judgment tests and assessment centres. Situational judgment tests present candidates with written scenarios and ask them to choose or rank possible responses. Assessment centres put candidates through behavioural simulations, role plays, and exercises observed and rated by trained assessors. Both are designed to assess how people would actually handle work-relevant situations, but they differ substantially in cost, complexity, and fidelity. The obvious question is which one works better. Shakeri and Lievens (2024) argue that this is the wrong question, and that asking it has obscured something more important.

The Problem With Prior Comparisons

Previous studies comparing the validity of situational judgment tests and assessment centres have typically done so by correlating each tool’s overall score with an overall rating of job performance. This approach, while convenient, conflates two different questions: how well does each tool predict performance in general, and how well does each tool predict specific performance dimensions it was designed to measure (Shakeri & Lievens, 2024)?

The problem is compounded by the fact that most prior comparisons have not kept the dimensions under investigation constant. If a situational judgment test is measuring interpersonal sensitivity while an assessment centre is measuring strategic thinking, comparing their overall validities tells you very little about the relative merits of the two methods for measuring the same thing. Shakeri and Lievens (2024) designed their study to address this directly, keeping the performance dimensions constant across both tools and examining validity at the dimension level rather than the overall score level.

The Study

Data were collected from 406 applicants for supervisory and management positions in a large Iranian steel industry company. The selection battery included a general mental ability test, a personality inventory, a situational judgment test, and an assessment centre. Supervisory ratings of job performance were collected across three specific dimensions: Thinking, which captured cognitive and analytical performance; Feeling, which captured interpersonal and emotional performance; and Power, which captured influence, authority, and leadership performance (Shakeri & Lievens, 2024). Both the situational judgment test and assessment centre were designed to assess all three dimensions, allowing a genuine head-to-head comparison.

What They Found

The results draw a clear and practically useful distinction between the two tools.

The assessment centre showed relatively high validity for all three performance dimensions, Thinking, Feeling, and Power, demonstrating broad predictive coverage across the range of constructs under investigation (Shakeri & Lievens, 2024).

The situational judgment test showed comparable validity to the assessment centre for the Thinking dimension, but was significantly weaker for the Feeling and Power dimensions (Shakeri & Lievens, 2024). In other words, when it came to predicting cognitively oriented performance, the two tools were roughly equivalent. When it came to predicting interpersonal and leadership performance, the assessment centre substantially outperformed the situational judgment test.

Incremental validity analyses confirmed this pattern. The situational judgment test added meaningful predictive value beyond the other tools for Thinking, but contributed considerably less for Feeling and Power. The assessment centre added meaningful value across all three dimensions (Shakeri & Lievens, 2024).

Why This Makes Conceptual Sense

The finding is theoretically coherent. Situational judgment tests present scenarios in written form and ask candidates to indicate what they would do. This format is well suited to assessing knowledge and judgment about cognitively demanding situations, where the key question is whether candidates understand what good thinking looks like in a given context. It is less well suited to assessing interpersonal and leadership qualities, which depend not just on knowing what to do but on the ability to actually do it under social pressure, with real emotional dynamics and real stakes. Assessment centres, with their behavioural simulations and direct observation of actual conduct, are better positioned to capture those qualities (Shakeri & Lievens, 2024).

This is essentially the distinction between knowing and doing, and it maps onto the longstanding concept of fidelity in simulation design. Higher fidelity simulations are more resource-intensive but capture aspects of performance that lower fidelity tools cannot reliably assess.

The Practical Implication

The authors are direct about what this means for practice. The question organisations should be asking is not which predictor has the best validity overall, but which predictor has the best validity for predicting what they specifically need to predict (Shakeri & Lievens, 2024). For roles where analytical and cognitive performance is the primary concern, a well-designed situational judgment test may be sufficient and offers significant cost advantages over an assessment centre. For roles where interpersonal, emotional, and leadership performance are central, an assessment centre is likely to be necessary, and a situational judgment test alone will leave important performance variance unassessed.

The strategic implication is that combining the two tools, rather than choosing between them, may offer the best of both: situational judgment tests for efficient assessment of cognitive and thinking dimensions, and assessment centres for the interpersonal and leadership dimensions that require higher fidelity observation.

Why It Matters

Selection system design has long been guided by overall validity coefficients, aggregated across dimensions and criteria. This study demonstrates that this level of analysis conceals meaningful variation in what different tools can and cannot measure. Organisations making decisions about which assessment methods to use on the basis of overall validity comparisons may be systematically underinvesting in the assessment of dimensions that matter for their specific roles, particularly the interpersonal and leadership qualities that are often most difficult to develop after hiring and most consequential for team and organisational performance.

Reference

Shakeri, I., & Lievens, F. (2024). A head-to-head comparison of situational judgment tests and assessment centers for measuring and predicting the same performance dimensions. International Journal of Selection and Assessment. Advance online publication. https://doi.org/10.1111/ijsa.12503