Test theory without true scores?

This paper traces the course of the consequences of viewing test responses as simply providing dichotomous data concerning ordinal relations. It begins by proposing that the score matrix is best considered to be items-plus-persons by items-plus-persons, and recording the wrongs as well as the rights. This shows how an underlying order is defined, and was used to provide the basis for a tailored testing procedure. It also was used to define a number of measures of test consistency. Test items provide person dominance relations, and the relations provided by one item can be in one of three relations with a second one: redundant, contradictory, or unique. Summary statistics concerning the number of relations of each kind are easy to get and provide useful information about the test, information which is related to but different from the usual statistics. These concepts can be extended to form the basis of a test theory which is based on ordinal statistics and frequency counts and which invokes the concept of true scores only in a limited sense.1979 Psychometric Society presidential address.I want to recognize the contributions which others have made to whatever I have accomplished. First to mention here are my teachers: principally Harold Gulliksen, Ledyard Tucker, and the late Edith Jay. Second, I would like to recognize the importance of my graduate students. Tom Reynolds has been especially important in developing the ideas that I will talk about here today, but at various times, the others have made major contributions in this and other topics. I would like to express also a debt to my family, primarily my wife, Rosemary, who herself has a longterm interest in the psychometric area. Finally, I must acknowledge the financial support of the NIMH some time ago, the Office of Naval Research, until about a year ago, and of the James McKeen Cattell Fund this past year.     

Precurrent self-prompting operants in children: "Remembering"

In Experiment I, one of three forms of collateral behavior was trained: Differential collateral behavior specific in form to one of two discriminative stimuli; Common collateral behavior of a single form regardless of the stimulus; or Nondifferential collateral behavior of either form regardless of the stimulus. Children were next given a short-delay matching-to-sample task in which the discriminative stimuli served as samples, and the children's previously trained collateral behavior terminated the delay and presented the comparison stimuli. Subjects engaging in sample-specific collateral behavior immediately acquired matching. Subjects engaging in sample-nonspecific collateral behavior failed to acquire matching or did so gradually. In Experiment II the minimal delay in the matching task was varied in a mixed sequence, first with collateral behavior required, and then with collateral behavior prohibited. When emitting collateral behavior Common and Nondifferential subjects showed delay-related decrements in matching while Differential subjects did not. When not emitting collateral behavior all subjects showed delay-related decrements in matching. Common and Nondifferential subjects matched more accurately when prohibited from emitting collateral behavior. Differential subjects matched more accurately when emitting collateral behavior. The results accord with Skinner's (1953, 1968) analysis of precurrent operants.     

A jackknife test of homogeneity of variance with paired replicates data

A nonparametric test of dispersion with paired replicates data is described which involves jackknifing logarithmic transformations of the ratio of variance estimates for the pre- and post-treatment populations. Results from a Monte Carlo simulation show that the test performs well underH _o and has good power properties. Examples are given of applying the procedure on psychiatric data.A referee called our attention to valuable references which related to our work, and he suggested the inclusion of competing jackknife procedures into our studies.     

Parameters affecting the maintenance of negatively reinforced key pecking

Three negative reinforcement experiments employing a key-peck response are described. In Experiment I, pigeons shocked on the average of twice per minute (imposed condition) could produce, by pecking a key, an alternate condition with correlated stimuli. Delayed shocks were added, across sessions, to the alternate condition until pecking stopped. Two of three pigeons continued to peck despite a 100% increase in shock frequency. In Experiment II, pigeons were shocked in the imposed condition four times per minute. The postresponse delay to shock was held constant by delivering, in the alternate condition, the next shock, or the next two, three, or four shocks from the imposed-condition shock schedule. All three subjects continued to peck with no change in delay to the first two postresponse shocks but with a 75% reduction in shock frequency. In Experiment III, a response produced an immediate shock followed by a shock-free period. Three of four subjects continued to respond despite reduced delay to shock. Delay-to-shock or shock-frequency reduction was sufficient to maintain key pecking, but neither was necessary. The conditions that negatively reinforce the pigeon's key peck were similar to conditions that negatively reinforce the rat's bar press.     

A parametric variation of delayed reinforcement in infants.

This study is an exploration of the parameters of delayed reinforcement with 6 infants (2 to 6 months old) in two experiments using single-subject repeated-reversal designs. In Experiment 1, unsignaled 3-s delayed reinforcement was used to increase infant vocalization rate when compared to a differential-reinforcement-of-other-than-vocalization condition and a yoked, no-contingency comparison condition. In Experiment 2, unsignaled 5-s delayed reinforcement was used to increase infant vocalization rate when compared to an alternating-treatments comparison condition. The alternating-treatments comparison consisted of 3-min components of differential reinforcement of other behavior and 3-min components of a nontreatment baseline. Successful conditioning was obtained in both experiments. These results contrast with those of previous infancy researchers who did not obtained conditioning with delays of 3 s and who attributed their findings to the limitations of the infant's memory capacity. We present an alternative conceptual framework and methodology for the analysis of delayed reinforcement in infants.     

Measurement invariance,factor analysis and factorial invariance

Several concepts are introduced and defined: measurement invariance, structural bias, weak measurement invariance, strong factorial invariance, and strict factorial invariance. It is shown that factorial invariance has implications for (weak) measurement invariance. Definitions of fairness in employment/admissions testing and salary equity are provided and it is argued that strict factorial invariance is required for fairness/equity to exist. Implications for item and test bias are developed and it is argued that item or test bias probably depends on the existence of latent variables that are irrelevant to the primary goal of test constructers.Presidential address delivered at the Annual Meeting of the Psychometric Society in Berkeley, California, June 18–20, 1993.     

信度系数与观测分数和潜在特质的相关比

在经典真分数模型中,信度系数R=D(T)/D(X)。通常认为,信度系数与项目反应理论没有什么联系。事实上,信度系数恰好等于考生观测分数与潜在特质分数的非线性相关比η_(xθ)~2=1-MD(x|θ)/D(X),据此我们得出估计信度系数的一种新途径,同时讨论了x、θ的相关系数与相关比η_(xθ)~2的关系。     

The role of body weight in resident-intruder aggression

The behavior of male NIH Swiss mice of various body weights in the resident-intruder test of aggression was investigated. Mice were housed individually for 10 days prior to the test, and allocated to six groups. In the first three groups body weights of residents and group-housed intruders were matched, and the animals were divided into light, average, or heavy groups. In the last three groups weights of the intruder mice were either matched with the residents, or intruders were lighter or heavier than the residents. We found that light residents spent significantly less time in aggressive behaviors and longer time in defensive behaviors than the other two groups. The heavy mice showed most social investigation. The body weights of intruders were also shown to affect the behavior in the test: those residents which had light opponents spent a longer time in aggressive behavior than those which had matched or heavy opponents. The resident mice with heavy opponents showed most defensive behaviors. To study whether pharmacological manipulation may have different effects on behavior in the resident-intruder test in mice having different weights, animals of light, average, and heavy body weight received a low dose (0.8 g/kg) of ethanol 30 min prior to the test. We did not note any effect of ethanol on aggressive behavior in the three groups. The results suggest that body weight plays a significant role in determining the level of aggression and defensive behaviors in the resident-intruder test.     

状元殿试对策的文化内涵

状元殿试对策有着深厚的文化底蕴。它既是对当时性文化现象的全面把握,又是对历时性文化内涵的深刻理解和灵活化用。状元殿试对策作为传统文化的重要载体之一,有着重要的文化研究价值,对当今考试文体也有着诸多启示和借鉴。     

Stepwise variable selection in factor analysis

It is very important to choose appropriate variables to be analyzed in multivariate analysis when there are many observed variables such as those in a questionnaire. What is actually done in scale construction with factor analysis is nothing but variable selection.In this paper, we take several goodness-of-fit statistics as measures of variable selection and develop backward elimination and forward selection procedures in exploratory factor analysis. Once factor analysis is done for a certain numberp of observed variables (thep-variable model is labeled the current model), simple formulas for predicted fit measures such as chi-square, GFI, CFI, IFI and RMSEA, developed in the field of the structural equation modeling, are provided for all models obtained by adding an external variable (so that the number of variables isp + 1) and for those by deleting an internal variable (so that the number isp – 1), provided that the number of factors is held constant.A programSEFA (Stepwise variable selection in Exploratory Factor Analysis) is developed to actually obtain a list of the fit measures for all such models. The list is very useful in determining which variable should be dropped from the current model to improve the fit of the current model. It is also useful in finding a suitable variable that may be added to the current model. A model with more appropriate variables makes more stable inference in general.The criteria traditionally often used for variable selection is magnitude of communalities. This criteria gives a different choice of variables and does not improve fit of the model in most cases.The URL of the programSEFA is http://koko15.hus.osaka-u.ac.jp/~harada/factor/stepwise/.