The EM algorithm for latent class analysis with equality constraints

The EM algorithm is a popular iterative method for estimating parameters in the latent class model where at each step the unknown parameters can be estimated simply as weighted sums of some latent proportions. The algorithm may also be used when some parameters are constrained to equal given constants or each other. It is shown that in the general case with equality constraints, the EM algorithm is not simple to apply because a nonlinear equation has to be solved. This problem arises, mainly, when equality constrints are defined over probabilities indifferent combinations of variables and latent classes. A simple condition is given in which, although probabilities in different variable-latent class combinations are constrained to be equal, the EM algorithm is still simple to apply.The authors are grateful to the Editor and the anonymous reviewers for their helpful comments on an earlier draft of this paper. C. C. Clogg and R. Luijkx are also acknowledged for verifying our results with their computer programs MLLSA and LCAG, respectively.     

Extensions of Rasch's multiplicative poisson model

Consideration will be given to a model developed by Rasch that assumes scores observed on some types of attainment tests can be regarded as realizations of a Poisson process. The parameter of the Poisson distribution is assumed to be a product of two other parameters, one pertaining to the ability of the subject and a second pertaining to the difficulty of the test. Rasch's model is expanded by assuming a prior distribution, with fixed but unknown parameters, for the subject parameters. The test parameters are considered fixed. Secondly, it will be shown how additional between- and within-subjects factors can be incorporated. Methods for testing the fit and estimating the parameters of the model will be discussed, and illustrated by empirical examples.     

Single-letter recognition accuracy and position information: Benefits with a symbolic cue

Summary This study reports two experiments designed to test a model that has as a basic assumption that position information is at the basis of the cueing benefits observed in single-item recognition tasks with accuracy as the dependent variable. In the experiments, two cueing conditions were used; a 100%-valid symbolic-cue condition and a neutral-cue condition. In Experiment 1, in which position information was sometimes missing, the symbolic cue brought a significant benefit. In Experiment 2, in which position information was always available, no benefit was obtained. These results are consistent with, and support, the model. Improved knowledge of position is at the basis of the cueing benefits observed in this type of task.     

Daily-life difficulties in person recognition reported by young and elderly subjects

Elderly people frequently complain spontaneously about their inability to remember people and their names. Naturalistic methods such as diaries, checklists, etc. provide useful means to study and make more explicit the nature of such memory difficulties among the elderly, as well as to better understand normal memory functioning. We developed a checklist (inspired by a study by Young, Hay and Ellis, 1985) to explore normal young and elderly people's difficulties in person recognition. The checklist is composed of four parts, each one corresponding to a particular context in which the difficulty took place. Each part has the same structure and consists of precise questions about the type of incident, the circumstances, the persons involved, and the way the incident ended. Three groups of normal subjects kept records of their difficulties and errors in recognizing people for 1 month: young subjects with a bad memory for faces, young subjects without particular problems of face memory, and a group of elderly subjects. A total of 299 records were collected. They were classified with respect to the functional components presumably implied in the process breakdown. The elderly subjects experienced difficulties with retrieving names, first names, or nicknames (Name Codes), while the young subjects with a bad memory for faces reported overall the greatest number of incidents and were particularly impaired in access to Face Recognition Units. A characterization of these two kinds of difficulties is proposed. Young subjects without problems of face memory presented equal numbers of difficulties at all stages of person recognition. These patterns of results will be discussed in terms of current cognitive models of person recognition.     

Interpersonal determinants of attitude change following counterattitudinal behavior

In a test of predictions derived from an identity-analytic model of self-presentational behavior, individuals who privately endorsed positive or negative attitudes about sexual behavior were asked to deliver a prosexuality speech while alone, while watched by observers, or while being watched by observers who questioned the morality of the subject’s actions. Subsequent attitude measures indicated that the subjects who initially adopted negative attitudes justified their behavior by expressing more favorable attitudes about sexuality, but only when no audience witnessed their speech. When an audience was present, these individuals emphasized their lack of choice. In contrast, subjects who privately endorsed positive attitudes publicly expressed less favorable attitudes when their morality was challenged by the observers. These findings suggest that attitude change following counterattitudinal behavior (a) stems from private image-maintenance needs as well as public self-presentational concerns, and (b) is sometimes designed to secure an image of morality as well as an image of consistency.     

Appraisal of behavioral measurement techniques for assessing dental anxiety and fear in children: A review

The aim of this article is to review and evaluate behavioral and physiological measurement techniques frequently used to assess dental anxiety and fear in children. Attention is given to the data collected, the empirical findings obtained, and the availability of normative data. The main focus, however, is on the reliability and validity. Results show that all questionnaires are open to criticism. Of the behavioral measures, Melamed's Behavior Profile Rating Scale is to be preferred to Frankl's Rating Scale, Venham Rating Scales, and Visual Analogue Scales. The main reasons are that Melamed's BPRS measures the behavior of the child more precisely and that it has superior psychometric properties. Furthermore, because of their practical, conceptual, and psychometric problems, physiological measures at this stage are found to be less appropriate for assessing dental fear in children. It is concluded that a behavioral measure is not always the ideal, but often the only available technique for assessing dental fear in children.