Estimation for the multiple factor model when data are missing

A maximum likelihood method of estimating the parameters of the multiple factor model when data are missing from the sample is presented. A Monte Carlo study compares the method with 5 heuristic methods of dealing with the problem. The present method shows some advantage in accuracy of estimation over the heuristic methods but is considerably more costly computationally.This paper is based on the author's doctoral dissertation at the Department of Psychology, University of Illinois at Urbana-Champaign. The author gratefully acknowledges the aid of Drs. Robert Bohrer, Charles Lewis, Robert Linn, Maurice Tatsuoka, and Ledyard Tucker.     

Ordinal data: An alternative distribution

To date, virtually all techniques appropriate for ordinal data are based on the uniform probability distribution over the permutations. In this paper we introduce and examine an alternative probability model for the distribution of ordinal data. Preliminary to deriving the expectations of Spearman's rho and Kendall's tau under this model, we show how to compute certain conditional expectations of rho and tau under the uniform distribution. The alternative probability model is then applied to ordinal test theory, and the calculation of true scores and test reliability are discussed.     

Recovery of structure in incomplete data by alscal

A Monte Carlo study was carried out in order to investigate the ability of ALSCAL to recover true structure inherent in simulated proximity measures when portions of the data are missing. All sets of simulated proximity measures were based on 30 stimuli and three dimensions, and selection of missing elements was done randomly. Properties of the simulated data varied according to (a) the number of individuals, (b) the level of random error, (c) the proportion of missing data, and (d) whether the same entries or different entries were deleted for each individual. Results showed that very accurate recovery of true distances, stimulus coordinates, and weight vectors could be achieved with as much as 60% missing data as long as sample size was sufficiently large and the level of random error was low.     

The principal components of mixed measurement level multivariate data: An alternating least squares method with optimal scaling features

A method is discussed which extends principal components analysis to the situation where the variables may be measured at a variety of scale levels (nominal, ordinal or interval), and where they may be either continuous or discrete. There are no restrictions on the mix of measurement characteristics and there may be any pattern of missing observations. The method scales the observations on each variable within the restrictions imposed by the variable's measurement characteristics, so that the deviation from the principal components model for a specified number of components is minimized in the least squares sense. An alternating least squares algorithm is discussed. An illustrative example is given.Copies of this paper and of the associated PRINCIPALS program may be obtained by writing to Forrest W. Young, Psychometric Laboratory, Davie Hall 013-A, Chapel Hill, NC 27514.     

Latent change classes in dichotomous data

Changes in dichotomous data caused by treatments can be analyzed by means of the so-called linear logistic model with relaxed assumptions (LLRA). The LLRA does not require observable criteria representing a single underlying latent trait, but it postulates the generalizability of the treatment effects over criteria and subjects. To test this latter crucial assumption, the mixture LLRA was proposed that allows directly unobservable types of subjects to have different treatment effects. As the earlier methods for estimating the parameters of the mixture LLRA have specific drawbacks, a further method based on the conditional maximum likelihood principle will be presented here. In contrast to the earlier conditional methods, it uses all of the dichotomous change data while having fewer parameters. Further, its goodness-of-fit tests become more sensitive to a falsely specified number of change-types even though the treatment effects are biased. For typically occurring small to moderate sample sizes, however, parametric bootstrapping of the distributions of the fit statistics is recommended for performing hypotheses tests. Finally, three applications of the new method to empirical data are described: first, about the effect of the so-called Trager psychophysical integration, second, about the effect of autogenic therapy on patients with psychosomatic symptoms, and, third, about the effect of religious education on the attitude towards sects. The mixture LLRA is implemented in the menu-driven program MIXLLRA which can be obtained from Ivo Ponocny via e-mail (ivo.ponocny@univie.ac.at).     

Validation of psychoanalytic theories: Towards a conceptualization of references

The authors discuss criteria for the validation of psychoanalytic theories and develop a heuristic and normative model of the references needed for this. Their core question in this paper is: can psychoanalytic theories be validated exclusively from within psychoanalytic theory (internal validation), or are references to sources of knowledge other than psychoanalysis also necessary (external validation)? They discuss aspects of the classic truth criteria correspondence and coherence, both from the point of view of contemporary psychoanalysis and of contemporary philosophy of science. The authors present arguments for both external and internal validation. Internal validation has to deal with the problems of subjectivity of observations and circularity of reasoning, external validation with the problem of relevance. They recommend a critical attitude towards psychoanalytic theories, which, by carefully scrutinizing weak points and invalidating observations in the theories, reduces the risk of wishful thinking. The authors conclude by sketching a heuristic model of validation. This model combines correspondence and coherence with internal and external validation into a four‐leaf model for references for the process of validating psychoanalytic theories.     

A comparison of normative data for the Trail Making Test from several countries: equivalence of norms and considerations for interpretation

The Trail Making Test may not be equivalent across cultures, i.e., differences in the scores across different cultures may not reveal real differences in the ability of the subjects on the construct being measured. In order to assess this hypothesis, normative samples from ten different countries were compared. Age decade subgroups across samples were ranked based on mean time taken to complete each part of the task. Large Z scores differences were found between these samples when comparing the first with the second, and the last in the rank. These differences were significant even when age and education were comparable across samples. Following Van de Vijver & Tanzer (1997), several possible sources of bias were identified. Incomparability of samples and administration differences were the most likely factors accounting for differences. Because of the lack of validity studies in the countries considered, no firm conclusions could be obtained regarding construct bias. Although the TMT may be measuring visual scanning, psychomotor speed and mental flexibility, normative data from different countries and cultures are not equivalent which might lead to serious diagnostic errors.     

From Zero to Sixty: Calibrating Real-Time Responses

Recent advances in data recording technology have given researchers new ways of collecting on-line and continuous data for analyzing input-output systems. For example, continuous response digital interfaces are increasingly used in psychophysics. The statistical problem related to these input-output systems reduces to linking time-varying covariates to a continuous response variate. Using real-time data obtained from an experiment in psychoacoustics, we showcase new statistical tools that incorporate dynamical elements of an input-output system. We employ functional data analysis (FDA) methods and a simple differential equation to analyze and model the continuous responses. Furthermore, we outline the issues involved in analyzing input-output systems when the exact form of the underlying mathematical model is not known. Finally, we develop a calibration method to facilitate inter-subject and intra-subject comparisons. This work was supported by grants from the National Sciences and Engineering Research Council (NSERC) to J. O. Ramsay and to D. J. Levitin, and by a grant from the Social Sciences and Humanities Research Council (SSHRC) to D. J. Levitin. We would like to thank Bennett Smith for designing and implementing the software used to conduct the pitch tracking experiment. Also, we wish to thank the research assistants in the Levitin Laboratory involved in the data collection: Catherine Chapados, Andrew Schaaf and Carla Himmelman. We would also like to acknowledge Giles Hooker’s work on implementing the generalized profiling software used within this paper.