Comparing Latent Structures of the Grade of Membership,Rasch, and Latent Class Models

This paper focuses on model interpretation issues and employs a geometric approach to compare the potential value of using the Grade of Membership (GoM) model in representing population heterogeneity. We consider population heterogeneity manifolds generated by letting subject specific parameters vary over their natural range, while keeping other population parameters fixed, in the marginal space (based on marginal probabilities) and in the full parameter space (based on cell probabilities). The case of a 2 × 2 contingency table is discussed in detail, and a generalization to 2^J tables with J ≥ 3 is sketched. Our approach highlights the main distinction between the GoM model and the probabilistic mixture of classes by demonstrating geometrically the difference between the concepts of partial and probabilistic memberships. By using the geometric approach we show that, in special cases, the GoM model can be thought of as being similar to an item response theory (IRT) model in representing population heterogeneity. Finally, we show that the GoM item parameters can provide quantities analogous to more general logistic IRT item parameters. As a latent structure model, the GoM model might be considered a useful alternative for a data analysis when both classes of extreme responses, and additional heterogeneity that cannot be captured by those latent classes, are expected in the population. This work was supported by Award #1R03 AG18986-01 from the National Institute on Aging and NIH grant #1R01 CA94212-01. The presentation of the ideas in this paper owes much to discussions with Stephen Fienberg and Brian Junker, Carnegie Mellon University. The author thanks Jim Ramsay and two anonymous reviewers for their valuable comments on earlier drafts of this paper.     

Every reinforcer counts: reinforcer magnitude and local preference

Six pigeons were trained on concurrent variable-interval schedules. Sessions consisted of seven components, each lasting 10 reinforcers, with the conditions of reinforcement differing between components. The component sequence was randomly selected without replacement. In Experiment 1, the concurrent-schedule reinforcer ratios in components were all equal to 1.0, but across components reinforcer-magnitude ratios varied from 1:7 through 7:1. Three different overall reinforcer rates were arranged across conditions. In Experiment 2, the reinforcer-rate ratios varied across components from 27:1 to 1:27, and the reinforcer-magnitude ratios for each alternative were changed across conditions from 1:7 to 7:1. The results of Experiment 1 replicated the results for changing reinforcer-rate ratios across components reported by Davison and Baum (2000, 2002): Sensitivity to reinforcer-magnitude ratios increased with increasing numbers of reinforcers in components. Sensitivity to magnitude ratio, however, fell short of sensitivity to reinforcer-rate ratio. The degree of carryover from component to component depended on the reinforcer rate. Larger reinforcers produced larger and longer postreinforcer preference pulses than did smaller reinforcers. Similar results were found in Experiment 2, except that sensitivity to reinforcer magnitude was considerably higher and was greater for magnitudes that differed more from one another. Visit durations following reinforcers measured either as number of responses emitted or time spent responding before a changeover were longer following larger than following smaller reinforcers, and were longer following sequences of same reinforcers than following other sequences. The results add to the growing body of research that informs model building at local levels.     

Concurrent ratio schedules: Fixed vs. variable response requirements

Rats were trained on concurrent fixed-ratio variable-ratio or concurrent fixed-ratio mixed-ratio schedules of food reinforcement. The variable-ratio schedule was composed of an arithmetic sequence of 11 ratios that averaged 50; the mixed-ratio schedule consisted of equiprobable ratios of 1 and 99. Fixed-ratio values, varied over experimental conditions, included 25, 35, 50, 60, and 99. The proportion of responses and time allocated to the variable- or mixed-ratio schedule increased as the size of the fixed ratio increased. For most subjects, higher proportions of responses and time were maintained on the fixed-ratio schedule at fixed-ratio values of 25 and 35; higher proportions of responses and time were maintained on the variable- or mixed-ratio schedule at fixed-ratio values of 50 or higher. On concurrent variable-ratio fixed-ratio schedules, the tendency for responding to be maintained exclusively by one schedule was related to the difference in local reinforcement rates obtained from those schedules. Exclusive responding was approximated when the difference in local reinforcement rates obtained from those schedules was large; responding was more evenly distributed between the schedules as the difference in the rates at which reinforcement was obtained from each decreased.     

Local response-rate constancy on concurrent variable-interval schedules of reinforcement

Concurrent variable-interval schedules were arranged with a main key that alternated in color and schedule assignment, along with a changeover key on which a small fixed ratio was required to changeover. Acceptable matching was observed with pigeons in two replications, but there was a tendency toward overmatching. Local response rates were found to differ for unequal schedules of a concurrent pair: local response rate was greater for the variable-interval schedule with the smaller average interreinforcement interval, but qualifications based on an interresponse-time analysis were discussed. In a second experiment, two 3-minute variable-interval schedules were arranged concurrently, and the experimental variable was the changeover procedure: either a changeover delay was incurred by each changeover or a small fixed ratio on a changeover key was required to complete a changeover. Changeover delays of 2 and 5 seconds were compared with a fixed-ratio changeover of five responses. The response output on the main key (associated with the variable-interval schedules) was greater when a changeover delay was arranged than when a fixed ratio was required to changeover. A detailed analysis of stripchart records showed that a 2-second delay generated an increased response rate for 3 seconds after a changeover, while the fixed-ratio requirement generated an increased rate during the first second only, followed by a depressed response rate for 2 seconds.     

When are item response models consistent with observed data?

The problem of deciding whether a set of mental test data is consistent with any one of a large class of item response models is considered. The classical assumption of locla independence is weakened to a new condition, local nonnegative dependence (LND). Necessary and sufficient conditions are derived for a LND item response model to fit a set of data. This leads to a condition that a set of data must satisfy if it is to be representable by any item response model that assumes both local independence and monotone item characteristic curves. An example is given to show that LND is strictly weaker than local independence. Thus rejection of LND models implies rejection of all item response models that assume local independence for a given set of data.This research was supported in part by Grant NIE-G-78-0157 to ETS from the NIE, by the Program Statistics Research Project, and by TOEFL Program Research. I would like to thank Dr. Douglas Jones of ETS for stimulating discussions during the early stages of this research, Dr. Frederick Lord of ETS for his encouragement of this work and comments on earlier drafts of this paper and Professor Robert Berk of Rutgers University for pointing out that conditions (a), (b) and (c) of Theorem 2 were also sufficient for LND and Monotonicity. Dr. Donald Alderman of ETS provided financial support for the development of a computer program to apply these results to data from the TOEFL program.     

Local homogeneity in latent trait models. A characterization of the homogeneous monotone irt model

The stochastic subject formulation of latent trait models contends that, within a given subject, the event of obtaining a certain response pattern may be probabilistic. Ordinary latent trait models do not imply that these within-subject probabilities are identical to the conditional probabilities specified by the model. The latter condition is called local homogeneity. It is shown that local homgeneity is equivalent to subpopulation invariance of the model. In case of the monotone IRT model, local homogeneity implies absence of item bias, absence of item specific traits, and the possibility to join overlapping subtests. The following characterization theorem is proved: the homogeneous monotone IRT model holds for a finite or countable item pool if and only if the pool is experimentally independent and pairwise nonnegative association holds in every positive subpopulation.This research was supported by the Dutch Interuniversity Graduate School of Psychometrics and Sociometrics. The authors wish to thank two reviewers for their thorough comments.     

高创造力个体的群体差异性研究

高创造力群体的界定是创造力研究的重要前提,但迄今极少有研究关注以往研究中高创造力群体间的同质性问题。本研究以创造力问题提出能力测验调查了两所高校（一所地方高校和一所教育部直属高校）的975名大学生的创造力,通过极端组方法（extreme groups approach, EGA）为大学生创造力水平分组,考察不同创造力分组之间的差别。研究发现：地方高校和部属高校的高创造力大学生被试,在创造力的流畅性、变通性和独特性三个维度上,均存在显著差异。本研究结果提示,大学生高创造力组间可能存在质的差别,因此相关研究在外推结果时需要采取谨慎态度;研究者在实验过程中和讨论文献结果的可比性时一定要关注样本的同质性问题。     

The Role of the Child in Adult Development

This paper provides theoretical and empirical support for the view that children play a large and often underappreciated role in adult development. Fifty parents and teachers were asked to discuss occasions when they learned something valuable from a child or when a child changed them in some significant way. Adults reported that children can cause them to (a) shift their values or priorities, (b) integrate memories or experiences previously disowned or repressed, (c) become more creative and cognitively flexible, or (d) look at the world with more wonder, awe, or curiosity. Werner's concept of genetic stratification (H. Werner, 1957) and Labouvie-Vief, Chiodo, Goguen, Diehl, and Orwoll's concept of dynamic intersubjectivity (G. Labouvie-Vief, L. M. Chiodo, L. A. Goquen, M. Diehl, & L. Orwoll, 1995) are used to explain the results.     

文本信息的激活与整合：阅读优生与差生的比较

探讨阅读优生与差生文本阅读过程中文本信息的激活与整合模式,包括3个实验。实验一通过运用不一致实验范式（inconsistent paradigm）考察阅读优生与阅读差生能否形成整体连贯性与局部连贯性。结果发现,阅读差生难以形成整体连贯性。实验二运用探测－核证范式（probe-verification paradigm）通过比较在三个位置后（人物特征描述、填充段落、目标句）的探测句的反应时间,探讨阅读差生不能形成整体连贯性的原因是由于不能激活先前的信息,还是整合信息时出现了问题。结果发现阅读差生与优生一样都可以激活先前的信息,说明阅读差生不能形成整体连贯性的原因可能是整合的问题。实验三进一步验证了阅读差生不能形成整体连贯性的原因是由于不能把已经激活的信息很好地加以整合所致。     

Comparing score tests and other local dependence diagnostics for the graded response model

Score tests for identifying locally dependent item pairs have been proposed for binary item response models. In this article, both the bifactor and the threshold shift score tests are generalized to the graded response model. For the bifactor test, the generalization is straightforward; it adds one secondary dimension associated only with one pair of items. For the threshold shift test, however, multiple generalizations are possible: in particular, conditional, uniform, and linear shift tests are discussed in this article. Simulation studies show that all of the score tests have accurate Type I error rates given large enough samples, although their small‐sample behaviour is not as good as that of Pearson's Χ² and M₂ as proposed in other studies for the purpose of local dependence (LD) detection. All score tests have the highest power to detect the LD which is consistent with their parametric form, and in this case they are uniformly more powerful than Χ² and M₂; even wrongly specified score tests are more powerful than Χ² and M₂ in most conditions. An example using empirical data is provided for illustration.