Dutch special forces operators, also known as commandos, perform in mentally and physically tough environments. An important question for recruitment and selection of commandos is whether they have particular personality traits. To answer this question, we first examined differences in personality traits between 110 experienced Dutch male commandos and a control sample of 275 men in the same age range. Second, we measured the personality traits at the start of the special forces selection program and compared the scores of candidates who later graduated (n = 53) or dropped out (n = 138). Multilevel Bayesian models and t tests revealed that commandos were less neurotic (d = −0.58), more conscientious (d = 0.45), and markedly less open to experiences (d = −1.13) than the matched civilian group. Furthermore, there was a tendency for graduates to be less neurotic (d = −0.27) and more conscientious (d = 0.24) than dropouts. For selection, personality traits do not appear discriminative enough for graduation success and other factors need to be accounted for as well, such as other psychological constructs and physical performance. On the other hand, these results provide interesting clues for using personality traits to recruit people for the special forces program.
No abstract
Neural networks can approximate any function given sufficiently many hidden units, which implies that they, in theory, can approximate human behavior. Recently, natural language processing has advanced rapidly due to increases in the amount of hidden units and in the size of the datasets. With these advances in natural language capabilities, we wondered whether state-of-the-art Large Language Models show human behavior. In this article, we demonstrate that these models show language comprehension and communication skills to solve problems, which are considered to be key features of human behavior. Moreover, the process by which such AI-based models encode information leads to errors which are also common in humans, such as being vulnerable to misleading questions, source amnesia, and being sensitive to small changes in wording. Given the similarities with human behavior, we discuss the potential applications of LLMs in social science research.We conclude that LLMs and their close alignment with human behavior may provide a valuable source of information that can be studied to gain a better understanding of human behavior.
Selecting the right individuals for a company, sports club, education program, or military unit has a large influence on the achievements of the organization. However, the approaches commonly used for selection are either not reporting predictive performance or are not interpretable (i.e., black box models). In this study, we introduce a novel approach to psychological research on selection, using both unexplainable and explainable machine learning models. We examined 239 recruits who performed a set of psychological tests (e.g., personality and intelligence) and physical tests (e.g., endurance and strength). On this data, we compared four machine learning models on their predictive performance, explainability, and stability. We found that a state-of-the-art stable rule-based (SIRUS) model was most suitable for classifying dropouts. With an averaged area under the curve score of 0.74, this model had a high predictive performance, and was most explainable and stable compared to the alternative models. Furthermore, we found that both psychological and physical variables were related to drop out. More specifically, a higher score on the 2800 meters time, sprint time, skin folds, sprint and agility, connectedness, fear of failure, and neuroticism were most strongly associated with drop out. Finally, we found that scoring higher on extraversion and amotivation was associated with drop out. We discuss how researchers and practitioners in the field of selection can benefit from these insights, and can add SIRUS to their selection toolbox.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.