Project Details
Projekt Print View

Taking the Situation in Situational Judgment Tests Seriously – A Situational Construal Perspective

Subject Area Social Psychology, Industrial and Organisational Psychology
Personality Psychology, Clinical and Medical Psychology, Methodology
Term from 2015 to 2023
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 271360450
 
Final Report Year 2023

Final Report Abstract

Situational Judgment Test (SJT) items typically present realistic situations along with several response options. For a long time, SJTs were considered as simulative instruments in personnel selection (akin to assessment centers). However, Krumm et al. (2015) identified that a substantial number of items from several SJTs were solvable despite their core element, the situation descriptions, being eliminated—thereby fundamentally challenging the view of SJTs as simulations. In response to this, Lievens and Motowidlo (2016) proposed to view SJTs as methods that capture general domain knowledge (and not so much situational judgment). However, several authors opposed this notion. This is where the two herein reported two funding periods set in. One of the objectives was to determine whether the context-dependency of SJTs affected their validity. We revealed that the validity of SJTs was only marginally affected by the availability of situation descriptions. As another objective, we sought to identify the SJT item components that drive SJT context-dependency. To this end, we took a multi-study approach and first developed SJTs to specifically include trait-related information. We then manipulated the availability of this information in the situation descriptions. Indeed, we found that the availability of trait-related situational information affected the context-dependency of SJTs. As another main objective, we took a closer look at the (situational) processes involved when responding to SJT items. We revealed that test takers subjective construal of the situation predicted SJT response behavior, but that this prediction held up even if the situational construal was made only on the basis of SJT response options. We also identified—across two independent data collections and multiple SJTs—that the Ability to Identify Criteria (ATIC), which was found to be relevant in other simulations, was not a substantial driver of SJT performance. A similar finding was made for situational strength: SJT items’ situational strength did not moderate the relationship between personality and SJT performance. We then deconstructed SJT items into all their components and examined whether social desirability influenced responses to SJT item components. We revealed that results varied substantially across SJT items, but overall full SJT items were best predicted by deconstructed SJT items, in which only response options were shown. Social desirability had a significant effect on only a few items. In a final study, we examined whether several types of situation perceptions differed across SJT item components and predicted responses to these item components. We found that situation perceptions were overall poor predictors of performance across different SJT item versions. Overall, findings from two consecutive research grants suggest that response behavior in SJTs is complex, considerably different from other simulations, not very “situational”, and likely mostly influenced by response options and—if included—by trait-relevant information.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung