等级效标分数的概率神经网络预测方法研究

针对基于统计学的等级效标分数预测存在的问题,提出了运用概率神经网络进行等级效标分数预测的方法。在20种条件下进行了计算机模拟实验,包括测验分数为单变量和多变量,以及各种水平的白噪声干扰条件,结果表明在测验分数为多变量的情况下,或者在有白噪声干扰的条件下,运用概率神经网络方法可以比统计学方法更好地对等级效标分数进行预测。     

Framing, intentions, and trust–choice incompatibility

The present paper examines how framing of messages and the intentions inferred from different—positive vs. negative—framings, interact with the development of trust. Empirical evidence is presented showing that different, logically equivalent, frames are supposedly interpreted as implying different intentions. Next, the relationship between different frames (and the corresponding intentions reflected from these frames) and trust are explored. Finally, and most important, the relationship between the assessment of trust, inferred from different frames, and the corresponding choice behavior resulting from these frames, are investigated. Specifically, consider agents A and B offering to sell exactly the same commodity, except that one agent formulates it in a positive and the other in a negative frame. The different frames may lead to different assessments of the trustworthiness of the two agents. Following common wisdom, if agent A is trusted more than B, then one should prefer to conduct transactions with the former rather than with the latter agent. Several experiments are presented that are incompatible with this conjecture. For example, when faced with a choice between two butchers, whose ground beef is advertised as containing 25% fat (negative frame) or 75% lean (positive frame), respectively, most people have more trust in the former yet most indicate they would buy their meat from the latter butcher. This phenomenon, in which negative framing weighs more in trust assessments, and positive framing weighs more in choice, is labeled trust–choice incompatibility. The robustness of the phenomenon is further demonstrated in several experiments, and possible explanations for its occurrence are discussed.     

Use and interpretation of test scores from limited cognitive test batteries: How g + Gc can equal g

Single scores from limited and unbalanced test batteries of cognitive ability can be ambiguous to interpret theoretically. In this study, a limited verbally and knowledge‐loaded cognitive test battery, from applicants to the Swedish police academies (N = 1,344), was examined to provide foundations for the use and interpretation of test scores. Three measurement models were compared: one single factor model and two bifactor models, which decomposed the variance of the battery into orthogonal components. The models were evaluated by fit indices and omega coefficients, and then applied to the prediction of academic performance. The overall prediction of all models was similar, although specific abilities also were found to provide substantial predictive validity over and above general intelligence (g). The findings provide support for the use of single scores in applied settings (selection), but suggest that it may be more appropriate to interpret such scores as composites of substantive components, and not just as measures of g.     

Empowering leadership,uncertainty avoidance,trust, and employee creativity: Interaction effects and a mediating mechanism

Taking an interactional approach, we hypothesized that (a) there is an interaction between empowering leadership, uncertainty avoidance, and trust that affects creativity, such that empowering leadership has the strongest positive relationship with creativity when the employees have high levels of uncertainty avoidance and trust their supervisors; and (b) creative self-efficacy mediates the effect that this three-way interaction between empowering leadership, uncertainty avoidance and trust has on creativity. In Study 1, we used a time-lagged research design, collecting multi-source data from 322 employees and their supervisors. The results of Study 1 supported our hypotheses. In Study 2, we used a more temporally rigorous research design in which data were collected in three stages, with a two-month time interval separating Stages 1 and 2, and Stages 2 and 3. On the basis of the time-lagged and multi-source data from 199 employees and their supervisors, Study 2 produced the same results as Study 1. We discuss the implications of these results for future research and practice.     

Longitudinal Effects of OCB on Cadre Selection and Pursuing a Career as Militia Cadre in the Swiss Armed Forces

In the present investigation longitudinal data obtained from recruits undergoing basic military training in the Swiss Armed Forces were used to examine the effects of organizational citizenship behavior (OCB) on a subsequent cadre selection and the willingness to voluntarily pursue a career as militia cadre. Results of the hierarchical logistic regression controlling for demographic variables, cadre recommendation, and allocation to the desired branch of service showed OCB predicted both the selection as cadre and the voluntariness to pursue a career as militia cadre. These effects show to what extent OCB may predict significant additional effort. Results are discussed with regard to further examination of the effects of OCB on career outcomes and selection decisions.     

Moderation of Stimulus Material on the Prediction of IQ with Infants' Performance in the Visual Expectation Paradigm: Do Greebles Make the Task More Challenging?

The objective of this study was to examine the role of the stimulus material for the prediction of later IQ by early learning measures in the Visual Expectation Paradigm (VExP). The VExP was assessed at 9 months using two types of stimuli, Greebles and human faces. Greebles were assumed to be associated with a higher load on working memory in comparison to human faces. IQ was assessed at 3 years and 4 months of age. Sixty‐four children participated at both assessments in this longitudinal study. The results show a moderation of the stimulus material on the prediction of IQ. With Greebles as stimulus material, a significant relationship between infants' learning and IQ was found. There was no relation when infants' learned in the VExP using human faces. Copyright © 2015 John Wiley & Sons, Ltd.     

Coordination and control of posture and ball release in basketball free-throw shooting

The objective of this study was to investigate the coordination of a whole-body task (basketball free-throw) in which success in performance outcome can be achieved through a manifold of combinations of postural and movement trajectory configurations. Participants were healthy men (19–24 years) with a range of skill levels that were tested for the accuracy of 50 basketball free-throws with both their dominant and non-dominant hand. The trial-to-trial variance in release parameters as well as postural stability of the shooter and synchronization of postural movement and ball release were strong predictors of performance, with non-elite shooters having a higher mean and variability of center-of-mass (COM) speed at the time of ball release. The synchronization between the time of peak COM and the time of ball release increased as a function of skill level and hand dominance, with the better performers releasing the ball more closely to the time of COM peak height. These findings reveal how, in addition to successfully controlling the trial-to-trial variability along the solution manifold of release parameters, the relative importance of the coordination of postural control and ball release properties on shooting success changes as a function of skill level.