Local optima in K-means clustering: what you don't know may hurt you

The popular K-means clustering method, as implemented in 3 commercial software packages (SPSS, SYSTAT, and SAS), generally provides solutions that are only locally optimal for a given set of data. Because none of these commercial implementations offer a reasonable mechanism to begin the K-means method at alternative starting points, separate routines were written within the MATLAB (Math-Works, 1999) environment that can be initialized randomly (these routines are provided at the end of the online version of this article in the PsycARTICLES database). Through the analysis of 2 empirical data sets and 810 simulated data sets, it is shown that the results provided by commercial packages are most likely locally optimal. These results suggest the need for some strategy to study the local optima problem for a specific data set or to identify methods for finding "good" starting values that might lead to the best solutions possible.     

Choice in a variable environment: effects of unequal reinforcer distributions

Six pigeons were trained in a procedure in which sessions included seven unsignaled components, each offering two pecking keys, and each providing a potentially different reinforcer ratio between the two keys. Across conditions, various combinations of reinforcer ratios and reinforcer-magnitude ratios were used to create unequal reinforcer distributions between the two alternatives when averaged across a session. The results extended previous research using the same basic procedure that had included only reinforcer distributions symmetrical around 1:1. Data analyses suggested that the variables controlling choice operated at a number of levels: First, individual reinforcers had local effects on choice; second, sequences of successive reinforcers obtained at the same alternative (continuations) had cumulative effects; and, third, when these sequences themselves occurred with greater frequency, their effects further cumulated. A reinforcer obtained at the other alternative following a sequence of continuations (a discontinuation) had a large effect and apparently reset choice to levels approximating the sessional reinforcer ratio.     

Correlations for the Vineland Adaptive Behavior Scales with Kaufman Brief Intelligence Test in a forensic sample

People with an intellectual disability are over-represented in the criminal justice system in many western countries. Identifying accused persons with intellectual disability is important if they are to receive protections available under the law. Accurate diagnosis is also relevant for correctional administrators, probation and parole services, and community services. Diagnosis of intellectual disability must be made on the basis of both cognitive skills (intelligence) and adaptive behavior. In this study, the Kaufman Brief Intelligence Test assessed intelligence, and the Vineland Adaptive Behavior Scales assessed adaptive behavior, through self-report. Tests were administered to 150 offenders, ranging in age from 13 to 53 years, in Australian prisons, juvenile detention centers, legal aid offices, and probation services. Pearson product-moment correlation coefficients calculated among all subtests and between total scores were significant. ROC curve analyses demonstrated that performance on each effectively predicted a standard score of less than 70 on the other one.     

Decision frame and procedural justice: interactive effects on perceptions of distributive justice

This work examined the effect of procedural justice and item frame on responses to positively and negatively worded survey items. Under conditions of low procedural justice, there is a significant difference in the rating of distributive justice items positively and negatively framed, but not when procedural justice is high. Implications for decision frame theory and biasing of responses to survey items are considered.     

SLOT: A research platform for investigating multimodal communication

In this article, we present the spatial logistics task (SLOT) platform for investigating multimodal communication between 2 human participants. Presented are the SLOT communication task and the software and hardware that has been developed to run SLOT experiments and record the participants’ multimodal behavior. SLOT offers a high level of flexibility in varying the context of the communication and is particularly useful in studies of the relationship between pen gestures and speech. We illustrate the use of the SLOT platform by discussing the results of some early experiments. The first is an experiment on negotiation with a one-way mirror between the participants, and the second is an exploratory study of automatic recognition of spontaneous pen gestures. The results of these studies demonstrate the usefulness of the SLOT platform for conducting multimodal communication research in both human-human and human-computer interactions.     

Telephone-administered cognitive behavior therapy for obsessive-compulsive disorder

Exposure with response prevention and cognitive behavior therapy are widely recognized as effective treatments for obsessive-compulsive disorder. Unfortunately, many people with obsessive-compulsive disorder--particularly those living in rural areas--do not have access to therapists providing these treatments. Accordingly, we investigated the efficacy of telephone-administered cognitive behavior therapy for obsessive-compulsive disorder. Two open trials are reported, for a total of 33 people with obsessive-compulsive disorder (without major depression). The first trial consisted of 12 weeks on a waiting list followed by 12 weeks of treatment (delayed treatment). The second trial consisted of 12 weeks of immediate treatment. Obsessive-compulsive symptoms did not change during the waiting period. Symptoms declined from pre- to post-treatment, with gains maintained at 12-week follow-up. For the pooled sample our pre-to-post-treatment effect size was as large or larger than those obtained in other studies of reduced contact treatment, and similar to those of face-to-face exposure with response prevention. Our proportion of treatment dropouts tended to be lower than those of other reduced contact interventions. The results suggest that telephone-administered cognitive behavior therapy is effective and well-tolerated, at least for people with obsessive-compulsive disorder without major depression. It remains to be seen whether this treatment is safe and effective when comorbid major depression is present.     

Sentence interference in the Stroop task

In two experiments on Stroop interference, we examined whether sentences can be processed without the intention of the reader. Participants named the ink colors in which words in sentences were printed, and the ink colors in which the same words, randomly arranged, were printed. In Experiment 1, sentences yielded longer response times (RTs) and more errors than did nonsentences, but only when they included words that were highly relevant to the color-naming task (i.e., color and color-related words). In Experiment 2, sentences yielded more errors than did nonsentences, and sentences in which the color words matched the set of ink colors yielded longer RTs than did nonsentences. The results indicate that sentence processing can be obligatory when the component words are highly relevant to the task.     

Expectancies About the Effects of Alcohol on the Self and on Others as Determinants of Alcohol Policy Attitudes1

The present study brings together 2 separate lines of research in the alcohol field: alcohol expectancies and policy attitudes. Expectancies concerning the short- and long-term effects of drinking (e.g., aggression), were predicted to be stronger determinants of support for alcohol availability control when the target person was another, compared to when discussing effects of one's own drinking. In a population-based survey conducted in 1992, 804 current drinkers were asked about self and other alcohol expectancies, as well as attitudes toward various alcohol policies. Results showed a distinct self-other discrepancy, with people more likely to expect alcohol to have a greater effect over others than over themselves. Further, it was found that these other-expectancies were among the strongest predictors of favoring tighter alcohol controls. These findings lend support to findings of a self-other bias in social psychology, and provide insight as to the structure of public opinion on alcohol policy.     

Brief psychodynamic counselling with young adults

Short-term counselling is frequently employed within university counselling services in which there can often be an inherent institutional objective of returning students to the academic treadmill as soon as possible (Coren, 1996). On the other hand, this is perhaps tempered by the availability of counselling throughout the student's academic career, affording them the opportunity to return for further intensive short-term work (or longer-term if indicated or required). However, research into the effectiveness of short-term counselling suggests that interventions of less than eight sessions are as effective as those lasting significantly longer (Barkham, 1989). This may be particularly true in the case of students experiencing transitory problems relating to their life at university. This paper uses a case illustration to examine the value of short-term psychodynamic counselling in a university situation.     

Understanding the Dynamics of Leadership: The Role of Follower Self-Concepts in the Leader/Follower Relationship

In this paper we integrate recent theory and research on the self-concept with leadership theory. We conceptualize the self as being defined at three levels: the individual, interpersonal, and group levels. The effects of the self on information processing and behavior are mediated by the working self-concept, which is composed of self-views, possible selves, and goals (standards). Leaders are seen as producing short-run changes by influencing the working self-concept and more enduring changes through the development of chronic schema. Subordinates are also seen as influencing leaders' self-schema, both individually and collectively. Both practical and theoretical implications of this perspective are discussed.