Identification of inconsistent variates in factor analysis

When some of observed variates do not conform to the model under consideration, they will have a serious effect on the results of statistical analysis. In factor analysis the model with inconsistent variates may result in improper solutions. In this article a useful method for identifying a variate as inconsistent is proposed in factor analysis. The procedure is based on the likelihood principle. Several statistical properties such as the effect of misspecified hypotheses, the problem of multiple comparisons, and robustness to violation of distributional assumptions are investigated. The procedure is illustrated by some examples.     

三种心理测量理论的信度观

目前,心理测量领域中主要存在三大理论派别。本文分别对这三种理论即经典测验理论、可概括性理论和项目反应理论作了简要介绍,着重分析这三种理论的信度观。文章讨论了这三种信度观的理论基础和研究方法,比较了它们的异同,指出经典测验理论存在的一些不足及概化理论和项目反应理论所作的改进。概化理论是对经典测验理论的扩展,它用多维的信度指标（概化系数）替代了经典测验理论的信度系数,项目反应理论则从信息量的角度出发,用项目信息函数、测验信息函数等指标更具体深入地反映项目、测验的测量可靠程度。     

Human Signal-detection Performance: Effects Of Signal Presentation Probabilities And Reinforcer Distributions

University students participated in one of four standard two-choice signal-detection experiments in which signal presentation probability was varied and the reinforcement distribution was held constant and equal. In Experiments 1, 3 and 4, subjects' performance showed a systematic response bias for reporting the stimulus presented least often. Experiments 1 and 4 showed that this effect was reliable with extended training and monetary, rather than point, reinforcement. In Experiment 2, all correct responses were signaled in some way, and this produced the opposite relationship between signal presentation probability and response bias. Experiments 1 and 3 found that explicitly deducting money (intended as punishment) for equal numbers of incorrect responses on each alternative, or varying the obtained overall rate of reinforcement, produced no clear change in response bias. The bias, shown by humans, for reporting the stimulus presented least often remains a challenge for theories of stimulus detection.     

Detecting a nonevent: Delayed presence-versus-absence discrimination in pigeons

Eight pigeons were trained on a delayed presence-versus-absence discrimination paradigm in which a sample stimulus was presented on some trials but not on others. If a sample was presented, then a response to one choice key produced food. If no sample was presented, a response to the other choice key produced food. The basic finding was that performance remained constant and well above 50% correct on no-sample trials as the retention interval increased, whereas performance dropped precipitously (to below 50% correct) on sample trials. In the second phase of the experiment, all of the trials were no-sample trials, and reinforcers were delivered probabilistically for one group of pigeons and according to time-based schedules for the other group. The exact reinforcement probabilities used in Phase 2 were those calculated to be in effect on no-sample trials in Phase 1 (according to a discrete-state model of performance). Subjects did not show exclusive preference for the richer alternative on no-sample trials in the first phase, but those in the probabilistic group developed near-exclusive preference for the richer alternative during the second phase. These data are inconsistent with the predictions of the discrete-state model, but are easily accommodated by an account based on signal detection theory, which also can be applied effectively to discrimination of event duration and the “subjective shortening” effect.     

Is Speech Arrest during Wada Testing a Valid Method for Determining Hemispheric Representation of Language?

Background and objective: The intracarotid amobarbital procedure, or Wada test, is the method of choice to determine hemispheric representation of language, and is routinely used in the presurgical evaluation for intractable epilepsy. Some investigators perform comprehensive language assessments, but others base language lateralization solely on speech arrest. This study sought to determine whether speech arrest alone during Wada testing provides valid data regarding language lateralization. Methods: The subjects (previously reported) were 21 patients evaluated for intractable epilepsy, who underwent language lateralization by Wada testing and functional MRI (FMRI). For each patient, language representation was determined by calculating: (1) a Wada laterality index based exclusively on speech arrest; (2) a Wada laterality index based on comprehensive language assessment; and (3) an FMRI laterality quotient. Correlation coefficients and categorical classifications were analyzed. Results: There was no significant correlation between the Wada laterality quotient derived from duration of speech arrest and either the comprehensive Wada language laterality score (r= .35,p= .12) or FMRI language laterality score (r= .32,p= .16). Categorical classification as left, right or bilateral language also showed marked discordance between speech arrest and the other two methods. Conclusion: Duration of speech arrest during Wada testing is not a valid measure of language dominance.     

词汇听觉使用频率在两类记忆测验中的比较研究

讨论了词语使用过程中视觉使用频率和听觉使用频率的区分。通过两个实验对听觉使用频率进行了探讨。实验一通过量表评定法得到了两组视觉使用频率相同,但听觉使用频率不同的双字词,实验二利用实验一得到的两组双字词对被试进行了记忆测验,在单一的听觉条件下,发现在自由回忆,听觉再认和模糊辨听这两类不同的记忆测验中出现了实验性分离现象;反映了听觉使用频率自身的一些特征。     

Rationally developed fake-good and fake-bad scales for the Jesness Inventory

Develops two validity scales for the Jesness Inventory using a rational approach: a fake-good scale, Lie (L); and a fake-bad scale, Overt Symptomatology (OS). Effectiveness was assessed using 293 male delinquents classified as fake-good, fake-bad, or honest based on a matched-pair MMPI-A. L was moderately effective in detecting the fake-good set, and OS tentatively effective in detecting the fake-bad set. Both correlated well with their MMPI-A counterparts. Sensitivity, specificity, positive predictive power, negative predictive power, and overall effectiveness data were reported. The L scale and OS scale were related to race but differences were less than 1 raw score point. L was unrelated to age. OS was related to age, with younger children showing more willingness to admit to symptomatology. Age-based modified T-score norms were developed for the newly constructed scales using 1142 male and 360 female delinquents. ages 13–18.     

Multiple-answer multiple-choice test items: Responding and scoring through bayes and minimax strategies

A multiple-answer multiple-choice test item has a certain number of alternatives,any number of which might be keyed. The examinee is also allowed to mark any number of alternatives. This increased flexibility over the one keyed alternative case is useful in practice but raises questions about appropriate scoring rules. In this article a certain class of item scoring rules called thebinary class is considered. The concepts ofstandard scoring rules and equivalence among these scoring rules are introduced in the misinformation model for which the traditional knowledge model is a special case. The examinee's strategy with respect to a scoring rule is examined. The critical role of a quantity called the scoring ratio is emphasized. In the case of examinee uncertainty about the number of correct alternatives on an item, a Bayes and a minimax strategy for the examinee are developed. Also an appropriate response for the examiner to the minimax strategy is outlined.Research partially supported under Grants N00014-67-A-0314-0022 from the Office of Naval Research and GS-32514 and MPS 75-07539 from the National Science Foundation.     

Reliability of multiple classifications

Cohen's kappa index is reformulated for multiple classifications based on exchangeable random variables. It is found that kappa is between 0 and 1 inclusive. Two characterizations for kappa are stated in terms of the relationship between such random variables. Within the normal test score model, kappa increases with test reliability and test length. Furthermore, when based on binary classifications, kappa is an inverse U-shaped function of the cutoff score. These trends also hold for the beta-binomial test score model.Paper read at the Spring meeting of the Psychometric Society, Bell Laboratories (Murray Hill, New Jersey), March 1976.The editorial assistance and helpful comments of Leonard S. Feldt, Sarah P. Seaman-Huynh, and the three referees are gratefully acknowledged.