Latent trait models and dichotomization of graded responses

This paper discusses thecompatibility of the polychotomous Rasch model with dichotomization of the response continuum. It is argued that in the case of graded responses, the response categories presented to the subject are essentially an arbitrary polychotomization of the response continuum, ranging for example from total rejection or disagreement to total acceptance or agreement of an item or statement. Because of this arbitrariness, the measurement outcome should be independent of the specific polychotomization applied, for example, presenting a specific multicategory response format should not affect the measurement outcome. When such is the case, the original polychotomous model is called compatible with dichotomization.A distinction is made between polychotomization or dichotomization before the fact, that is, in constructing the response format, and polycho- or dichotomization after the fact, for example in dichotomizing existing graded response data.It is shown that, at least in case of dichotomization after-the-fact, the polychotomous Rasch model is not compatible with dichotomization, unless a rather special condition of the model parameters is met. Insofar as it may be argued that dichotomization before the fact is not essentially different from dichotomization after the fact, the value of the unidimensional polychotomous Rasch model is consequently questionable. The impact of our conclusion on related models is also discussed.     

Linear Analogue Self-Assessment and Procrustean Measurement: A Critical Review of Visual Analogue Scaling in Pain Assessment

Visual analogue scales (VASs) are one of the most widely used self-report measures of clinical pain. This article reviews the empirical literature on linear analogue self-assessment (LASA) and critically examines the features that appear to have made it such an attractive measurement option in pain assessment. It is concluded that analogue scaling does not withstand critical scrutiny as a primary measure of either pain intensity or pain affect and that, in most circumstances, the overall clinical utility of VASs is inferior to that of more structured pain indices.     

Human d-amphetamine drug discrimination: methamphetamine and hydromorphone.

Standard measures of subjective and discriminative effects of drugs were compared in 5 human volunteers. Subjects responded on a second-order color-tracking procedure, where 30 mg of d-amphetamine served as a discriminative stimulus for one response and its absence as the discriminative stimulus for another response. Self-reported subjective effects were assessed concurrently using the single-dose questionnaire, subscales of the Addiction Research Center Inventory, and several analogue rating scales. On different days following discrimination acquisition, varying doses of d-amphetamine, methamphetamine, and hydromorphone were administered. In these test sessions, either response was reinforced. Methamphetamine and d-amphetamine occasioned dose-related increases in d-amphetamine appropriate responding; hydromorphone did not. Methamphetamine and d-amphetamine occasioned dose-related increases in reports of the drug received being most like "speed"; hydromorphone occasioned dose-related increases in reports of the drug received being most like "dope." All three drugs occasioned dose-related increases in reports of drug liking, and increases in the morphine-benzedrine group, amphetamine, and benzedrine group scales of the Addiction Research Center Inventory. This experiment demonstrated that although explicit discriminative control of behavior by a drug may covary with drug identification, it does not necessarily covary with other self-reported subjective effects. Thus, the complementary nature of the data provided by drug discrimination and standard subjective-effects measures provides quantitative and qualitative data useful in studying both relatively novel compounds and the behavioral biology of psychoactive drugs in general.     

The Revised Conners' Parent Rating Scale (CPRS-R): Factor Structure, Reliability, and Criterion Validity

The Conners' Parent Rating Scale (CPRS) is a popular research and clinical tool for obtaining parental reports of childhood behavior problems. The present study introduces a revised CPRS (CPRS-R) which has norms derived from a large, representative sample of North American children, uses confirmatory factor analysis to develop a definitive factor structure, and has an updated item content to reflect recent knowledge and developments concerning childhood behavior problems. Exploratory and confirmatory factor-analytic results revealed a seven-factor model including the following factors: Cognitive Problems, Oppositional, Hyperactivity-Impulsivity, Anxious-Shy, Perfectionism, Social Problems, and Psychosomatic. The psychometric properties of the revised scale appear adequate as demonstrated by good internal reliability coefficients, high test-retest reliability, and effective discriminatory power. Advantages of the CPRS-R include a corresponding factor structure with the Conners' Teacher Rating Scale—Revised and comprehensive symptom coverage for attention deficit hyperactivity disorder (ADHD) and related disorders. Factor congruence with the original CPRS as well as similarities with other parent rating scales are discussed.     

Effects of diagnostic classification systems on clinical hypothesis-generation

In this study it was determined whether (a) classification as opposed to absence of classification has an effect on the quality of clinical hypotheses (b) the DSM-III-R and the CBCL have a different effect on the quality of clinical hypotheses, and (c) the potential difference between the DSM-III-R and the CBCL is moderated by the different number of syndromes identified by these systems. To investigate these questions, an experiment was conducted in which 86 clinicians generated hypotheses for six cases. The clinicians were divided into a DSM-III-R, a CBCL, and a control group. Of the six cases, two were classified by both classification systems as one syndrome, two were classified as one syndrome by the DSM-III-R but as two syndromes by the CBCL, and two were classified as two syndromes by the DSM-III-R but as one syndrome by the CBCL. The quality of the hypotheses was determined by means of four dependent variables selected from an overview of qualitative criteria: explanatory value, redundancy, possibility of operationalization, and specificity. No differences between the CBCL and the control groups were found. The DSM group performed better than the control group regarding explanatory value and redundancy. The DSM-III-R group also scored better than the CBCL group regarding explanatory value, particularly when the number of identified syndromes was two for the CBCL and one for the DSM-III-R.     

Validation of the Children's Assertive Behavior Scale

Relationships between sociometric indices and scores of 160 elementary-school children on the Children's Assertive Behavior Scale (CABS) were examined to determine whether this instrument differentiates children of varying levels of social competence. Children who obtained either high or low scores on peer nominations of friendship and admiration were identified and compared on CABS passive, aggressive, and assertive scores. Children who scored high on positive peer nominations responded in a significantly less aggressive manner than children who received low scores on such nominations. These effects were observed for the CABS format that consisted of stimulus situations involving adults but not for the form involving peers. With regard to responses toward peers, a significant interaction between sex and sociometric status emerged. These findings suggest the importance of incorporating subject variables such as race and sex as well as assessments of behavior toward both adults and children in isolating and training components of social skill.The authors wish to express their appreciation to Jean Birbilis, Gail Hammersly, Marieta Knopf, and Donna Wadley for their assistance in data collection and analysis.     

Reduction of sadistic sexual arousal by olfactory aversion: a case study

A sexual sadist was treated by an olfactory aversion procedure for eight weeks. Pre-treatment assessment of sexual arousal showed high levels of response to sadistic stimuli as well as high arousal to non-sadistic heterosexual stimuli. In treatment the inhalation of the gas of an odoriferous chemical agent, valeric acid, was paired with slide presentations of sadistic materials while penile erection was monitored. The aversive stimulus immediately and permanently suppressed the response for the duration of treatment. In separate measurement sessions, with valeric acid absent, deviant response remained suppressed and nondeviant response was unaffected. Post-treatment reassessment revealed very low levels of response to sadistic stimuli. Follow-up sessions showed that the deviant response was absent eight months following the conclusion of treatment.     

Attention demands of movements as a function of their duration and velocity

The attention demands of initiating and controlling discrete movements were examined as a function of their movement time (MT) and average movement velocity. Experiment 1 showed that the attention required to execute a movement decreased as MT decreased, although Experiment 2 through independently manipulating MT and movement velocity, revealed that movement velocity is the key determiner of attention demands rather than MT. The attention demands of preparing a high velocity movement are greater than during its execution with the reverse being the case for relatively slow velocity movements. The results are compatible with the view that it is the initiation of error corrections that are attention demanding (Keele 1973).