Elaboration et validation de tests de jugement situationnel comme outil de sélection professionnelle : guide méthodologique

Current views in personnel selection recognize the necessity for situational judgements tests (SJT) which are closed to the real work context of companies. The SJTs have also become a popular selection tool across Europe. SJTs are attractive because they show good validity, positive applicants reactions and can be cost effective to test large group of applicants at once. The aim of this paper is to describe a structured methodology for the elaboration of such test. Each stage is illustrated by examples and psychometrics validation issues are discussed.     

How Long Do Offenders Escape Arrest? Using DNA Traces to Analyse when Serial Offenders Are Caught

Why is it that some serial offenders are arrested quickly and others only after a long period, or never at all? What characterises offenders who continue to escape arrest despite their continued involvement in crime? To be able to answer these questions, arrested (identified) offenders must be compared with never arrested (unidentified) offenders. In this paper, data from the Dutch DNA database are used to assess which characteristics of the criminal behaviour of unidentified offenders influence the probability that they will subsequently be arrested. DNA data offer a unique opportunity to compare crime series of identified offenders with crime series of yet unidentified offenders. Using the Cox proportional hazards model, we tested whether the number of crimes committed, offence specialization, and offence seriousness affect the probability of arrest of serial offenders. Results showed that as an offender commits more crimes, the probability that he will be arrested increases and that offence specialization decreases the probability of arrest. Another conclusion drawn is that DNA traces offer unique opportunities for criminological research. We discuss the limitations of this new data source and make suggestions for future research using DNA traces and for future research that might improve the current study. Copyright © 2012 John Wiley & Sons, Ltd.     

Books Received

Leg-to-body ratio (LBR) is one of the morphological traits that influences a person's attractiveness. To date, studies confirming that hypothesis have been conducted mainly in Western cultures. They have shown that the average or slightly higher-than-the-average LBR is perceived to be attractive in women. In the case of men, results were more ambiguous; however generally shorter or similar LBRs compared to females were attractive. Here, data on LBR preferences of a traditional, semi-nomad ethnic group (i.e., the Himba of northern Namibia, n = 81) are reported. Also in Himba people LBR influences a person's attractiveness. Similar to Western societies, extremely high and low LBRs were unattractive. However, contrary to previous findings, Himba preferred women of relatively low LBR and men of relatively high LBR.     

EFFECTS OF A SIGNALED DELAY TO REINFORCEMENT IN THE PREVIOUS AND UPCOMING RATIOS ON BETWEEN-RATIO PAUSING IN FIXED-RATIO SCHEDULES

Domestic hens responded under multiple fixed‐ratio fixed‐ratio schedules with equal fixed ratios. One component provided immediate reinforcement and the other provided reinforcement after a delay, signaled by the offset of the key light. The components were presented quasirandomly so that all four possible transitions occurred in each session. The delay was varied over 0, 4, 8, 16, and 32 s with fixed‐ratio 5 schedules, and over 0, 8 and 32 s with fixed‐ratio 1, 15 and 40 schedules. Main effects of fixed‐ratio value and delay duration were detected on between‐ratio pauses. Pauses were longer when the multiple‐schedule stimulus correlated with a delayed‐reinforcer component was presented, with the longest pauses occurring at the transition from a component with an immediate reinforcer to one with a delayed reinforcer. Pause durations were shortest during immediate components. Overall, both the presence or absence of a delay in the upcoming component, and the presence or absence of a delay in the preceding component affected pause length, but the upcoming delay had the larger effect. Thus changes in delay had similar effects to past reports of the effects of changes in response force, response requirement, and reinforcer magnitude in multiple fixed‐ratio fixed‐ratio schedules.     

Multilevel Reliability Measures of Latent Scores Within an Item Response Theory Framework

Abstract

This paper evaluated multilevel reliability measures in two-level nested designs (e.g., students nested within teachers) within an item response theory framework. A simulation study was implemented to investigate the behavior of the multilevel reliability measures and the uncertainty associated with the measures in various multilevel designs regarding the number of clusters, cluster sizes, and intraclass correlations (ICCs), and in different test lengths, for two parameterizations of multilevel item response models with separate item discriminations or the same item discrimination over levels. Marginal maximum likelihood estimation (MMLE)-multiple imputation and Bayesian analysis were employed to evaluate the accuracy of the multilevel reliability measures and the empirical coverage rates of Monte Carlo (MC) confidence or credible intervals. Considering the accuracy of the multilevel reliability measures and the empirical coverage rate of the intervals, the results lead us to generally recommend MMLE-multiple imputation. In the model with separate item discriminations over levels, marginally acceptable accuracy of the multilevel reliability measures and empirical coverage rate of the MC confidence intervals were found in a limited condition, 200 clusters, 30 cluster size, .2 ICC, and 40 items, in MMLE-multiple imputation. In the model with the same item discrimination over levels, the accuracy of the multilevel reliability measures and the empirical coverage rate of the MC confidence intervals were acceptable in all multilevel designs we considered with 40 items under MMLE-multiple imputation. We discuss these findings and provide guidelines for reporting multilevel reliability measures.     

基于作答时间数据的改变点分析在检测加速作答中的探索——已知和未知项目参数

相对于传统的离散作答数据, 作答时间作为连续数据, 可以提供更多信息。改变点分析(change point analysis)技术在心理和教育领域是一个比较新的技术。本文一方面对改变点分析在心理测量领域的应用进行了一个综合的总结和分析; 另一方面, 将基于作答数据的两种改变点分析统计量推广到作答时间数据, 将改变点分析技术应用到测验异常作答模式：加速作答speededness的检测上。采用两种检验方法：似然比检验和Wald检验, 分别在已知和未知项目参数的条件下, 实现异常作答模式的检测。结果表明, 所采用的方法对于加速作答行为的检测具有很高的检验力, 同时能够很好的控制I类错误率。实证数据分析进一步表明本文中所使用的方法具有应用价值。     

新世纪20年国内假设检验及其关联问题的方法学研究

新世纪20年来国内假设检验方法学研究内容可分为如下几类: 零假设显著性检验的不足、p值的使用问题、心理学研究的可重复性问题、效应量、检验力、等效性检验、其他与假设检验关联的研究。零假设显著性检验已经发展成一套组合流程: 为了保证检验力和节省成本, 实验研究需要做先验检验力分析预估样本容量, 但问卷超过160人在传统统计中就没有必要这样做。当拒绝零假设时, 应当结合效应量做出结论。当不拒绝零假设时, 需要报告后验检验力; 如果效应量中或大而检验力不够高, 则可增加被试再行分析, 但这一过程应主动披露, 报告最后的实际p值并对可能犯的第一类错误率做出评估。     

计算机化分类测验终止规则的类别、特点及应用

计算机化分类测验(Computerized Classification Testing, CCT)能够高效地对被试进行分类, 已广泛应用于合格性测验及临床心理学中。作为CCT的重要组成部分, 终止规则决定测验何时停止以及将被试最终划分到何种类别, 因此直接影响测验效率及分类准确率。已有的三大类终止规则(似然比规则、贝叶斯决策理论规则及置信区间规则)的核心思想分别为构造假设检验、设计损失函数和比较置信区间相对位置。同时, 在不同测验情境下, CCT的终止规则发展出不同的具体形式。未来研究可以继续开发贝叶斯规则、考虑多维多类别情境以及结合作答时间和机器学习算法。针对测验实际需求, 三类终止规则在合格性测验上均有应用潜力, 而临床问卷则倾向应用贝叶斯规则。     

If I imagine it,then it happened: The Implicit Truth Value of imaginary representations

Imagination sometimes leads people to behave, feel, and think as though imagined events were real even when they know they were not. In this paper, we suggest that some understanding of these phenomena can be achieved by differentiating between Implicit Truth Value (ITV), a spontaneous truth evaluation, and Explicit Truth Value (ETV), a self-reported truth judgment. In three experiments, we measure ITV using the autobiographical Implicit Association Test (Sartori, Agosta, Zogmaister, Ferrara, & Castiello, 2008), which has been used to assess which of two autobiographical events is true. Our findings demonstrate that imagining an event, like experiencing an event, increases its ITV, even when people explicitly acknowledge the imagined event as false (Experiments 1a and 1b). Furthermore, we show that imagined representations generated from a first-person perspective have higher ITV than imagined representations generated from a third-person perspective (Experiment 2). Our findings suggest that implicit and explicit measures of truth differ in their sensitivity to properties underlying truth judgment. We discuss the contribution of characterizing events according to both ITV and ETV to the understanding of various psychological phenomena, such as lying and self-deception.