Test theory without an answer key

A general model is presented for homogeneous, dichotomous items when the answer key is not known a priori. The model is structurally related to the two-class latent structure model with the roles of respondents and items interchanged. For very small sets of respondents, iterative maximum likelihood estimates of the parameters can be obtained by existing methods. For other situations, new estimation methods are developed and assessed with Monte Carlo data. The answer key can be accurately reconstructed with relatively small sets of respondents. The model is useful when a researcher wants to study objectively the knowledge possessed by members of a culturally coherent group that the researcher is not a member of.This research was supported by NSF Grant No. SES-8320173 to the authors. We gratefully acknowledge comments and suggestions from John Boyd, Tarow Indow, and Kathy Maher as well as the editor and several anonymous referees.     

Above suspicion: Cognitive and intentional aspects of the ability to lie

This paper looks at the attribution of the ability to lie and not at lying or lies. It also departs from more familiar approaches by focussing on the appraisal of an ability and not on the ability in itself. We believe that this attribution perspective is required to bring out the cognitive and intentional basis of the ability to lie.     

An automated test development of parallel tests from a seed test

Binary programming models are presented to generate parallel tests from an itembank. The parallel tests are created to match item for item an existing seed test and match user supplied taxonomic specifications. The taxonomic specifications may be either obtained from the seed test or from some other user requirement. An algorithm is presented along with computational results to indicate the overall efficiency of the process. Empirical findings based on an itembank for the Arithmetic Reasoning section of the Armed Services Vocational Aptitude Battery are given.The Office of Naval Research, Program in Cognitive Science, N00014-87-C-0696 partially supported the work of Douglas H. Jones. The Rutgers Research Resource Committee of the Graduate School of Management partially supported the work of Douglas H. Jones and Ing-Long Wu. A Thomas and Betts research fellowship partially supported the work of Ing-Long Wu. The Human Resources Laboratory, United States Air Force, partially supported the work of Ronald Armstrong. The authors benefited from conversations with Dr. Wayne Shore, Operational Technologies, San Antonio, Texas. The order of authors' names is alphabetical and denotes equal authorship.     

Ordinal consistency and ordinal true scores

This paper argues that test data are ordinal, that latent trait scores are only determined ordinally, and that test data are used largely for ordinal purposes. Therefore it is desirable to develop a test theory based only on ordinal assumptions. A set of ordinal assumptions is presented, including an ordinal version of local independence. From these assumptions it is first shown that the gamma-correlation between two tests is the product of their gamma-correlations with the true latent order. The theory is generalized to allow for heterogeneous tests by defining a weighted average local independence. The tau-correlations between total score and the latent order can be found in both homogeneous and heterogeneous cases, and a system of differential item weighting to maximize the tau-correlation between weighted items and the latent order is provided. Thus a purely ordinal test theory seems possible.Part of this work was done while the author was a Visiting Fellow at Macquarrie University. The paper has benefitted from discussions with Professors Thomas J. Reynolds and Roderick P. McDonald and from the comments of several anonymous reviewers.     

Effects of relative reinforcer frequency on complex color detection

Pigeons were trained under a discrete-trials detection procedure in which one of a set of color stimuli was presented on the center key and a single response turned off the stimulus and illuminated two side keys. Single responses to one or the other side key produced occasional reinforcers depending on the value of the color stimulus. In Experiment 1, one color-stimulus set comprised 559, 564, 569, and 574 nm, and right-key pecks were occasionally reinforced following presentations of members of this set. The other stimulus set comprised 579, 584, 589, and 594 nm, and left-key pecks were occasionally reinforced following presentations of members of this set. Across seven experimental conditions, the left/(left + right) relative reinforcer frequency was varied from .1 to .9. In Experiment 2, one stimulus set contained only one member, 574 nm, and right-key responses were occasionally reinforced following its presentation. Over 12 experimental conditions, two manipulations were carried out. First, the number of stimuli comprising the other stimulus set was increased from one (579 nm) to two (579 and 584 nm) to three (579, 584, and 589 nm) and to four (579, 584, 589, and 594 nm), and left-key responses were reinforced occasionally following center-key presentations of members of this set. Second, for each stimulus combination, the left/(left + right) relative reinforcer frequency was varied from .1 to .5 to .9 across three experimental conditions. The principal finding of Experiments 1 and 2 was that reinforcers and stimuli interacted in their effects on behavior. In Experiment 3, pairs of adjacent stimuli (5 nm apart) in the range 559 to 594 nm were presented in each experimental condition, and the left/(left + right) relative reinforcer frequency was held constant at .5. The data from all three experiments were analyzed according to a detection model describing performance in multiple-stimulus two-response procedures. This model provided independent measures of stimulus discriminability, contingency discriminability, and bias. The analysis showed that (a) consistent with the color-naming function, pigeons were better able to discriminate between higher nanometer values than lower nanometer values; (b) their ability to discriminate between the stimuli was independent of the number of wavelengths comprising each stimulus set; (c) they allocated delivered reinforcers very accurately to the previously emitted response; and (d) no consistent biases emerged.     

Multifactor measurement of bodily feelings: Conceptual development,scale construction,and empirical validation

A series of studies is described demonstrating a coherent exploratory and confirmatory factor-analytic approach to the scaling of self-reported bodily feelings. This approach attempts to optimize the orthogonality, parsimony, and meaningfulness of a multifactor inventory. Psychometric properties of the resulting 35-item Bodily Feeling Scale (BFS) are presented and its comparability with other inventories is discussed. A validity study using item-content-based scales of the Minnesota Multiphasic Personality Inventory showed both concurrent and discriminative validity of the BFS. Implications for clinical and research utility are discussed.Portions of this research were conducted while the first author was supported by a University Graduate Fellowship awarded by the Graduate School of Arts and Sciences at Vanderbilt University.Professor Nunnally passed away in August 1982. His major contributions to the several stages of this research program are recognized through coauthorship.     

The interaction of stimulus and reinforcer control in complex temporal discrimination.

Six pigeons were trained in a discrete-trials signal-detection procedure to discriminate between a fixed-duration stimulus (5 s or 20 s) and a set of variable durations ranging from 2.5 s to 57.5 s in steps of 5 s. For each fixed-duration stimulus, the ratio of reinforcer frequencies contingent upon reporting the fixed versus the variable stimulus was systematically manipulated. Detection performance was well controlled by both the stimulus value and the reinforcer ratio. Both the discriminability between the fixed duration and the set of variable durations, and the discriminability between the fixed duration and each of the variable durations, were independent of the reinforcer-frequency ratio when discriminability was measured as log d. The sensitivity of response bias to reinforcement-ratio changes was independent of the value of the fixed duration, but was not independent of the discriminability of the variable durations from the fixed durations. Under current models, discriminability measures in complex temporal discrimination may be independent of biasing manipulations, but bias measures are not independent of stimulus values.     

A nonparametric approach for assessing latent trait unidimensionality

Assuming a nonparametric family of item response theory models, a theory-based procedure for testing the hypothesis of unidimensionality of the latent space is proposed. The asymptotic distribution of the test statistic is derived assuming unidimensionality, thereby establishing an asymptotically valid statistical test of the unidimensionality of the latent trait. Based upon a new notion of dimensionality, the test is shown to have asymptotic power 1. A 6300 trial Monte Carlo study using published item parameter estimates of widely used standardized tests indicates conservative adherence to the nominal level of significance and statistical power averaging 81 out of 100 rejections for examinee sample sizes and psychological test lengths often incurred in practice.The referees' comments were remarkably detailed and greatly enhanced the writeup and sensitized the author to certain pertinent issues. Discussions with Fritz Drasgow, Lloyd Humphreys, Dennis Jennings, Brian Junker, Robert Linn, Ratna Nandakumar, and Robin Shealy were also very useful.This research was supported by the Office of Naval Research under grant N00014-84-K-0186; NR 150-533, and by the National Science Foundation under grant DMS 85-03321.     

A taxonomy of item response models

A number of models for categorical item response data have been proposed in recent years. The models appear to be quite different. However, they may usefully be organized as members of only three distinct classes, within which the models are distinguished only by assumptions and constraints on their parameters. “Difference models” are appropriate for ordered responses, “divide-by-total” models may be used for either ordered or nominal responses, and “left-side added” models are used for multiple-choice responses with guessing. The details of the taxonomy and the models are described in this paper. The present study was supported in part by two postdoctoral fellowships awarded to Lynne Steinberg: an Educational Testing Service Postdoctoral Fellowship at ETS, Princeton, NJ and an NIMH Individual National Research Service Award at Stanford University, Stanford, CA. Helpful comments by the editor and three anonymous reviewers are gratefully acknowledged.