On the robustness of maximum likelihood scaling for violations of the error model

A Monte Carlo study was conducted to investigate the robustness of the assumed error distribution in maximum likelihood estimation models for multidimensional scaling. Data sets generated according to the lognormal, the normal, and the rectangular distribution were analysed with the log-normal error model in Ramsay's MULTISCALE program package. The results show that violations of the assumed error distribution have virtually no effect on the estimated distance parameters. In a comparison among several dimensionality tests, the corrected version of thex ² test, as proposed by Ramsay, yielded the best results, and turned out to be quite robust against violations of the error model.     

Probabilistic multidimensional scaling: Complete and incomplete data

Simple procedures are described for obtaining maximum likelihood estimates of the location and uncertainty parameters of the Hefner model. This model is a probabilistic, multidimensional scaling model, which assigns a multivariate normal distribution to each stimulus point. It is shown that for such a model, standard nonmetric and metric algorithms are not appropriate. A procedure is also described for constructing incomplete data sets, by taking into consideration the degree of familiarity the subject has for each stimulus. Maximum likelihood estimates are developed both for complete and incomplete data sets. This research was supported by National Science Grant No. SOC76-20517. The first author would especially like to express his gratitude to the Netherlands Institute for Advanced Study for its very substantial help with this research.     

Optimal appropriateness measurement

The test-taking behavior of some examinees may be so idiosyncratic that their test scores may not be comparable to the scores of more typical examinees. Appropriateness measurement attempts to use answer patterns to recognize atypical examinees. In this report appropriateness measurement procedures are viewed as statistical tests for choosing between a null hypothesis of normal test-taking behavior and an alternative hypothesis of atypical test-taking behavior. Most powerful tests for inappropriateness are described together with methods for computing their power. A recursion greatly simplifying the calculation of optimal test statistics is described and illustrated.The work reported in this article was supported by United States Office of Naval Research contracts N00014-79C-0752, NR 154-445 and N00014-83K-0397, NR 150-518, Michael V. Levine, Principal Investigator.     

A comparison of the efficiency and accuracy of BILOG and LOGIST

Comparisons are made between BILOG version 2.2 and LOGIST 5.0 Version 2.5 in estimating the item parameters, traits, item characteristic functions (ICFs), and test characteristic functions (TCFs) for the three-parameter logistic model. Data analyzed are simulated item responses for 1000 simulees and one 10-item test, four 20-item tests, and four 40-item tests. LOGIST usually was faster than BILOG in producing maximum likelihood estimates. BILOG almost always produced more accurate estimates of individual item parameters. In estimating ICFs and TCFs BILOG was more accurate for the 10-item test, and the two programs were about equally accurate for the 20- and 40-item tests.I am grateful to Robert J. Mislevy, Martha L. Stocking, and Marilyn S. Wingersky for many helpful comments on an earlier version of this paper. I would also like to thank Hamid Kamrani and Bongmyoung Park for getting LOGIST and BILOG running and keeping them running under changing computer systems at CTB/McGraw-Hill.     

A multidimensional scaling model for the size-weight illusion

The kinds of individual differences in perceptions permitted by the weighted euclidean model for multidimensional scaling (e.g., INDSCAL) are much more restricted than those allowed by Tucker's Three-mode Multidimensional Scaling (TMMDS) model or Carroll's Idiosyncratic Scaling (IDIOSCAL) model. Although, in some situations the more general models would seem desirable, investigators have been reluctant to use them because they are subject to transformational indeterminacies which complicate interpretation. In this article, we show how these indeterminacies can be removed by constructing specific models of the phenomenon under investigation. As an example of this approach, a model of the size-weight illusion is developed and applied to data from two experiments, with highly meaningful results. The same data are also analyzed using INDSCAL. Of the two solutions, only the one obtained by using the size-weight model allows examination of individual differences in the strength of the illusion; INDSCAL can not represent such differences. In this sample, however, individual differences in illusion strength turn out to be minor. Hence the INDSCAL solution, while less informative than the size-weight solution, is nonetheless easily interpretable.This paper is based on the first author's doctoral dissertation at the Department of Psychology, University of Illinois at Urbana-Champaign. The aid of Professor Ledyard R Tucker is gratefully acknowledged.     

Capturing the noonday demon: development and validation of the State Boredom Inventory

For centuries, scholars have positioned state boredom as an impediment to organizational productivity and performance given its unpleasant and distracting qualities. However, research on state boredom has been impeded by a lack of definitional consensus and measurement issues. In this article, we sought to advance organizational research on state boredom by developing the State Boredom Inventory (SBI), an 11-item measure grounded in a theoretically derived definition of state boredom. Across five studies using 10 independent samples, we develop the SBI and provide validity evidence for our measure, including content validity and convergent and discriminant validity. Our data support the conceptualization of state boredom as a higher-order multidimensional construct with three underlying dimensions: disengagement, unpleasant low arousal, and inattention.     

Construct and concurrent validity of the Cambridge neuropsychological automated tests in Portuguese older adults without neuropsychiatric diagnoses and with Alzheimer’s disease dementia

We aimed to analyze the construct and concurrent validity of the Rapid Visual Information Processing (RVP), Paired Associates Learning (PAL), Reaction Time (RTI), and Spatial Working Memory (SWM) tests from the Cambridge Neuropsychological Test Automated Battery (CANTAB®). Inclusion criteria were checked in a first session. The CANTAB and additional pencil-and-paper tests were administered within 1 week. The participants (aged 69–96 years) were 137 Portuguese adults without neuropsychiatric diagnoses and 37 adults with mild-to-moderate Alzheimer’s disease dementia. Comparisons were made between the CANTAB tests and between these tests and the Rey Complex Figure Test (RCFT), Verbal Fluency (VF) test, and some Wechsler Memory Scale-III and Wechsler Adult Intelligence Scale-III subtests. Most intra-test correlations were stronger than the CANTAB inter-test correlations. The RVP correlated more with VF animals (.44), the PAL with RCFT immediate recall (–.52), the RTI with RVP mean latency (.42), and the SWM with Spatial Span backward (–.39).     

Comprehensive driving assessment: neuropsychological testing and on-road evaluation of brain injured patients

The study investigates the correspondence between neuropsychological test results and on-road driving performance among 55 patients with a CT-verified brain damage or documented neurological disorder (cerebrovascular accident: 43, traumatic brain injury: 5, multiple sclerosis: 4, other: 3). 5 patients showed unimpaired test profiles and passed the on-road evaluation. 18 patients showed severe neuropsychological deficits contrary to driving and were not recommended for on-road evaluation. Of the remaining 32 patients with some neuropsychological deficits, all 100% in the minor impaired group (n = 8) passed the driving evaluation, compared to 69% in the mildly impaired (n = 16) and 38% in the moderately impaired group (n = 8). Measures of reduced visuoconstructive ability, reaction time, visual attention, and awareness of cognitive impairments, were found to discriminate between groups. It is concluded that neuropsychological assessment of targeted functions provide an ecological valid prediction of driving skill after brain damage, but that on-road evaluation is needed as supplement in cases with ambiguous test findings.