The asymptotic posterior normality of the latent trait in an IRT model

It has long been part of the item response theory (IRT) folklore that under the usual empirical Bayes unidimensional IRT modeling approach, the posterior distribution of examinee ability given test response is approximately normal for a long test. Under very general and nonrestrictive nonparametric assumptions, we make this claim rigorous for a broad class of latent models.This research was partially supported by Office of Naval Research Cognitive and Neural Sciences Grant N0014-J-90-1940, 442-1548, National Science Foundation Mathematics Grant NSF-DMS-91-01436, and the National Center for Supercomputing Applications. We wish to thank Kumar Joag-dev and Zhiliang Ying for enlightening suggestions concerning the proof of the basic result.The authors wish to thank Kumar Joag-Dev, Brian Junker, Bert Green, Paul Holland, Robert Mislevy, and especially Zhiliang Ying for their useful comments and discussions.     

Alternative response training, differential reinforcement of other behavior, and extinction in squirrel monkeys (Saimiri sciureus)

In Experiment I, (a) extinction, (b) extinction plus reinforcement of a discrete alternative response, and (c) differential reinforcement of other behavior were each correlated with a different stimulus in a three-component multiple schedule. The alternative-response procedure more rapidly and completely suppressed behavior than did differential reinforcement of other behavior. Differential reinforcement of other behavior was slightly more effective than extinction alone. In Experiment II, reinforcement of specific alternative behavior during extinction and differential reinforcement of other behavior were used in two components, while one component continued to provide reinforcement for the original response. Once again, the alternative-response procedure was most effective in reducing responding as long as it remained in effect. However, the responding partially recovered when reinforcement for competing behavior was discontinued. In general, responding was less readily reduced by differential reinforcement of other behavior than by the specific alternative-response procedure.     

Positive reinforcement and the elimination of reinforced responses

Key pecking was maintained on a fixed-interval schedule while either a differential-reinforcement-of-not-responding or a fixed-time schedule was imposed simultaneously. The lower the time parameter of the not-responding schedule, the lower was the response rate. Similar effects occurred with the fixed-time schedule, if the pigeons had experience with reinforcement for not responding. Otherwise the effects were less orderly, to the extent that rate could reach maximum with the lowest-valued fixed-time schedule. The not-responding and the response-independent schedules had similar effects on rate in experienced pigeons only when the time parameter or nominal frequency of food presentation was considered. When considered in terms of obtained frequency of food presentation, reinforcement of not responding produced larger decrements in rate than did the fixed-time schedule.     

THE EFFECTIVENESS OF REINFORCEMENT,RESPONSE COST,AND MIXED PROGRAMS ON CLASSROOM BEHAVIORS1

Six elementary school children served as subjects in an experiment comparing the relative effectiveness of (1) token reinforcement, in which children received tokens for attending and for correct arithmetic performance; (2) response cost, in which children received “free” tokens at the start of a period but lost them for inattention and for arithmetic performance below a specified level; and (3) a combination of both token reinforcement and response cost. During training, the six subjects received all three procedures in counterbalanced sequence. The effects of the three procedures were assessed by a within-subject comparison divided into three phases: (i) baseline, (ii) training, (iii) withdrawal of tokens. Introduction of the three token procedures markedly increased the two dependent measures. However, there were no differences across the procedures in the amount of change produced in either attending behavior or arithmetic performance. During baseline, the subjects averaged 29% attending behavior and 6.4 correct problems. These levels increased to 85% for attending behavior and 11.4 correct problems for arithmetic performance during training. Removal of all token procedures significantly decreased attending behavior (to an average of 65%), but produced a nonsignificant reduction in arithmetic performance (to an average of 7.6 correct problems). There was evidence that this lack of differential effects of the three token procedures was not due to an inability to discriminate among them. Furthermore, the subjects were evenly divided in their preference for the three procedures.     

Force and rate relations in responding during variable-interval reinforcement

Four rats responded on one-minute variable-interval schedules with several variations in peak-force of response required for food reinforcement. Measures of peak force and rate were taken for the responses, which were the downward exertions of force against a static force-transducing operandum. The analysis distinguished responses, a generic class of measured behavior, from criterion responses, an operationally specified subclass required for reinforcement. Absolute rate of response showed no systematic change, but the rate of responses meeting a newly required criterion of peak-force invariably increased through changes in the absolute rate of response, the relative-frequency distributions of peak force, or some combination of both. The relative frequency of responses meeting an elevated force criterion during variable-interval reinforcement exceeded that maintained with the same criterion with continuous reinforcement. The requirement of more effortful responding for reinforcement does not necessarily reduce response rate. Conformity of the behavior to the requirement for reinforcement is the salient effect.     

Self-reinforcement

Self-reinforcement in operant situations generally refers to those arrangements in which the subject delivers to himself a consequence, contingent on his behavior. However, it is noted that the definition of all other types of reinforcement make its delivery contingent on the subject's behavior. What is actually at issue is the agent who defines whether or not the response required for reinforcement has been met. In self-reinforcement, the subject himself defines this. In the laboratory, this requirement is machine-defined; in school examinations, it is teacher-defined; and in many clinical self-control situations, it is also independently defined. A reinforcement contingency presupposes such independence, absent in self-reinforcement. Implications for research and practice are discussed and alternative formulations are offered.     

Evolution and relationship maintenance: Fertility cues lead committed men to devalue relationship alternatives

An evolutionary perspective provides a conceptual framework with which to study adaptive processes that facilitate the maintenance of long-term relationships. One common relationship threat people face involves exposure to tempting relationship alternatives. Desirable relationship alternatives can threaten people's relationship commitment and, consequently, evoke relationship maintenance processes designed to protect people's relationship esteem. The current research examined whether cues of female ovulation–an important reproductive variable–evokes relationship maintenance processes in committed men. Men rated the attractiveness of a normally cycling relationship alternative at various points in her menstrual cycle. Unlike single men, who rated the woman as especially attractive when she was highly fertile, committed men exhibited the opposite pattern, rating her as less attractive during her period of peak fertility. This research illustrates the utility of integrating theories of relationship maintenance with evolutionary psychological theories of mating.     

The interaction between stimulus and reinforcer control on remembering.

In a symbolic matching-to-sample task, 6 pigeons obtained food by pecking a red side key when the brighter of two white lights had been presented on the center key and by pecking a green side key when the dimmer of two white lights had been presented on the center key. Across Part 1 and Parts 6 to 10, the delay between sample-stimulus presentation and the availability of the choice keys was varied between 0 s and 25 s. Across Parts 1 to 5, the delay between the emission of a correct choice and the delivery of a reinforcer was varied between 0 s and 30 s. Although increasing both types of delay decreased stimulus discriminability, lengthening the stimulus-choice delay produced a greater decrement in choice accuracy than did lengthening the choice-reinforcer delay. Additionally, the relative reinforcer rate for correct choice was varied across both types of delay. The sensitivity of behavior to the distribution of reinforcers decreased as discriminability decreased under both procedures. These data are consistent with the view, based on the generalized matching law, that sample stimuli and reinforcers interact in their control over remembering.     

Using Twitter to Teach Reader‐Oriented Biblical Interpretation: “Tweading” the Gospel of Mark

Twitter offers an engaging way to introduce students to reader‐oriented interpretation of the Bible. The exercise described here introduces students to the idea that the reader has a role in the production of a text's meaning, which thus varies from reader to reader. Twitter enables us to capture the real‐time thoughts of a variety of respondents to the text of Mark as it is read aloud. Students can concretely observe the effects of particular textual moments on individual respondents as well as analyze their general interpretive stances with regard to the text as a whole. Students come to grasp that the meaning of the text varies depending on the reader, setting the stage for more complex theoretical discussion of reader‐response theory, the reader's role in the production of meaning, the adjudication of “allowed” and “disallowed” interpretations, and the appropriateness of “reader‐response” criticisms for texts that were composed to be encountered orally.     

Putting the “IRT” in “Dirty”: Item response theory analyses of the Dark Triad Dirty Dozen—An efficient measure of narcissism,psychopathy, and Machiavellianism

The Dark Triad Dirty Dozen is a new, concise personality inventory designed to measure individual differences in narcissism, psychopathy, and Machiavellianism in sub-clinical populations. To date, the psychometric work on this measure has used classical test theory to assess its validity, reliability, and structure. In two independent samples (N = 1014), we provide the first analyses of this measure using item response theory. In both samples, the Dark Triad Dirty Dozen efficiently recovered information and showed some evidence of differential scale functioning by sex, with a lower endorsement threshold for men compared to women. Results shed new light on the measurement of these “dark” personality traits.