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.     

Sample size determinations for the two rater kappa statistic

This paper gives a method for determining a sample size that will achieve a prespecified bound on confidence interval width for the interrater agreement measure,. The same results can be used when a prespecified power is desired for testing hypotheses about the value of kappa. An example from the literature is used to illustrate the methods proposed here.     

Involutions defined by monadic terms

We prove that there are two involutions defined by monadic terms that characterize Monadic Algebras. We further prove that the variety of Monadic Algebras is the smallest variety of Interior Algebras where these involutions give rise to an interpretation from the variety of Bounded Distributive Lattices into it.Research funded by Fondecyt, project N 764-1987.     

A representation theorem for languages with generalized quantifiers through back-and-forth methods

We obtain in this paper a representation of the formulae of extensions ofL by generalized quantifiers through functors between categories of first-order structures and partial isomorphisms. The main tool in the proofs is the back-and-forth technique. As a corollary we obtain the Caicedo's version of Fraïssés theorem characterizing elementary equivalence for such languages. We also discuss informally some geometrical interpretations of our results.     

Closed-class words as first syllables do interfere with lexical decisions for nonwords: implications for theories of agrammatism

It has been proposed that a principal cause of the agrammatism of some Broca's aphasics is that such patients, unlike normal subjects, are unable to make use of a special retrieval mechanism for closed-class ("function") words (D. C. Bradley, 1978, Computational distinctions of vocabulary type, Unpublished Ph.D. thesis; D. C. Bradley, M. F. Garrett, & E. B. Zurif, 1980. In D. Caplan (Ed.), Biological studies of mental processes). The main evidence for the existence of such a mechanism consisted of two observations: (1) the recognition of open-class words was observed to be frequency-sensitive, but that of closed-class words was not; and (2) lexical decisions for nonwords which began with open-class words were delayed, whereas there was no such interference for nonwords which began with closed-class words. However, the first of these observations has proved nonreplicable (e.g., B. Gordon & A. Caramazza, 1982, Brain and Language, 15, 143-160, 1983, Brain and Language, 19, 335-345; J. Segui, J. Mehler, W. Frauenfelder, & J. Morton, 1982, Neuropsychologia, 20, 615-627), and in the present paper, three lexical decision experiments are reported in which it is found that, when certain confounding variables are controlled, nonwords which begin with closed-class words are subject to interference. Moreover, contrary to a suggestion of Kolk and Blomert (1985, Brain and Language, 26, 94-105) the interference is independent of the presence of closed-class items in the lexical decision word list. It seems, then, that closed-class words are not qualitatively different from open-class words with respect either to frequency sensitivity or to nonword interference, and in consequence, the above proposed explanation of agrammatism is left without major empirical support.     

Role of the area postrema in three putative measures of motion sickness in the rat

After thermal cauterization of the area postrema in rats the absence of conditioned taste aversion to sucrose paired with lithium chloride (0.15 M, 3.3 ml/kg) was used as a pharmacologic/behavioral index of area postrema damage. In a subsequent experiment the effects of area postrema lesions on three measures proposed as species-relevant measures of motion sickness were studied, using off-vertical rotation at 150 degrees/s for either 30 or 90 min. Lesions of area postrema did not alter postrotational suppression of drinking or amount of defecation during motion. The initial acquisition of conditioned taste aversion to a novel cider vinegar solution paired with motion was not affected by lesioning of the area postrema, but these taste aversions extinguished more slowly in lesioned rats than in sham-operates or intact controls. Results are discussed in terms of proposed humoral factors which may induce motion sickness and in light of recent data on the role of the area postrema in similar measures in species possessing the complete emetic reflex.     

Understanding the central processing limit in consistent-mapping visual search tasks

Effects of load (i.e., the number of stimuli in the display) have been observed in multiple-frame studies using a consistent mapping of stimuli to responses (e.g., Fisher, 1982, 1984). In a series of four experiments, it is shown that these effects are not the consequence of differences across the high- and low-load conditions in either decision noise or peripheral masking. Additionally, it is shown that of two modes of limited capacity (a limited-channel and divided-capacity model) considered as possible explanations of load effects in tasks where subjects are required to locate a target, only one--the limited-channel model--is consistent with the results from all three location tasks. Finally, it is argued that the limited-channel model predicts not only the behavior observed in the four consistent-mapping experiments reported in this article but also the behavior observed in several related consistent-mapping tasks (Kleiss & Lane, 1986; Shiffrin & Gardner, 1972).