Anthony J. Conger

Integration and generalization of kappas for multiple raters.

J. A. Cohens kappa (1960) for measuring agreement between 2 raters, using a nominal scale, has been extended for use with multiple raters by R. J. Light (1971) and J. L. Fleiss (1971). In the present article, these indices are analyzed and reformulated in terms of agreement statistics based on all

Kappa Reliabilities for Continuous Behaviors and Events

Cohens kappa for measuring agreement between two observers using a discrete nominal scale is extended to measuring agreement over time for continuous nominal scales. The continuous kappa coefficient avoids problems encountered by the arbitrary division of real time durations into presence/absence frequencies in discrete intervals. The extension is simple but issues of independence and number of observations pose problems for significance testing.

Injury severity and outcome: a meta-analysis of prospective studies on TBI outcome.

OBJECTIVE The recent movement to apply evidenced-based approaches to medical and rehabilitation care has increased the importance of approximating outcomes as early in the recovery process as possible. The relationship between injury severity and outcome following traumatic brain injury (TBI), however, has remained unclear due to the variety of predictor and criterion variables used throughout the literature. METHOD A meta-analysis of eligible prospective studies that assessed the bivariate association between injury severity and outcome at 1-year postinjury was conducted. RESULTS Twenty-six studies met the inclusion criteria (total N = 21,050 patients). Injury severity was a significant predictor of outcome at 1-year postinjury (r = .257). Homogeneity testing by means of the Q test, Q(n), indicated that injury severity measurement, Q(68) = 1140.76, p < .00001, outcome measurement, Q(42) = 516.63, p < .00001, and outcome measurement construct, Q(4) = 14.65, p = .006, were significant moderators of the injury severity/outcome link. Further, the magnitude of the effect between injury severity and outcome resulted from a significant interaction between the measure of injury severity and the outcome construct. CONCLUSIONS Overall, measures of injury severity were most significantly associated with measures of global outcome and most poorly associated with measures of satisfaction with life. Additionally, a significant interaction was found between the measure of injury severity and the outcome construct indicating that different measures of injury severity more precisely predict one outcome construct over another. Methodological concerns were discussed and recommendations for creating a more parsimonious and integrated literature base were made.

Mood and forbidden foods’ influence on perceptions of binge eating

This study consists of two experiments investigating the effects of induced mood and food type on perceptions of eating in imagined and real eating situations. A total of 212 female undergraduates representing the continuum of bulimic symptomatology were induced with either elated or depressed moods using a standardized mood-induction procedure. They were then either asked to imagine themselves in a situation with either forbidden or non-forbidden foods (Experiment 1) or else were presented with a buffet of forbidden or non-forbidden foods and asked to eat (Experiment 2). Participants subsequently reported their perception of their eating behavior (i.e., amount of control, meal rating: from a snack to a binge; and meal feeling: from great to bad). Results revealed limited support for affect regulation models of bulimia nervosa when the participants consumed food, but no support for the theory when they imagined eating. Conversely, forbidden foods were found to influence perceptions in the imagined eating situation, but not when the participants ate. Implications of these results are discussed.

Comparability of selection instruments in studies of heterosexual-social problem behaviors

To evaluate the comparability of the selection methods used in analogue studies on heterosexual-social problem behaviors, 67 undergraduate males were administered four commonly used paper-and-pencil self-report instruments designed to assess both “dating experience” and “social anxiety”. Major findings from the analyses were (a) the majority of the intercorrelations were statistically significant, although most were relatively low in magnitude, (b) based on a principal component analysis, the battery of instruments appeared multidimensional in nature with considerable content and method variance present among instruments, and (c) using traditional selection criteria, all instruments produced relatively independent subject samples, and little was gained from knowing classification on one instrument in predicting classification on another. Implications for past and future research in this analogue research area and the current clinical research on social skills training were discussed, along with the lack of convergence of these instruments.

What can the WISC-R measure?

The WISC-R was investigated by using measures of profile (multivariate) reliability in order to determine its most reliable dimensions and the precision and similarity of the multivariate structure across age groups. Due to differences among the 11 age groups in both subscale reliabilities and true score covariance matrices, it was concluded that the precision of measurement differed across age groups. This finding was further supported by a comparison of canonical reliability coefficients and composites computed for each age group. However, exhaustive analyses of Varimax rotated profile dimensions indicated that the structure of the WISC-R subscales is rather stable across age groups, but the reliability of that structure differs systematically. A synthesis of the analyses indicated that (1) the WISC-R allows highly reliable comparisons of profile levels (Full-Scale IQ) at each age level; that (2) reasonably reliable comparisons of Verbal-Performance differences can be made at each age level; but that (3) for other comparisons, caution should be exercised because of age group differences and potentially high unreliability. Two strategies for the interpretation of WISC-R profiles, which take into account the above findings, are offered.

Maximally Reliable Composites for Unidimensional Measures

A general approach to obtaining weights and reliability coefficients of maximally reliable composites is offered for a variety of test theoretic models which assume a single common underlying dimension. The reliability maximizing weights are related to the theoretically specified true score scaling weights to show there is a constant relationship between them that is invariant under separate linear transformations on each variable in the system. Furthermore, an argument is made that test theoretic relations should be derived for the most general model available and not for unnecessarily constrained models.

The Relationship of Anger and Social Skills to Psychological Symptoms

The link between anger, social skills, and psychological symptoms was investigated in a college population. Seven hundred and nine individuals were administered the State Trait Anger Expression Inventory, the Anger Inventory, the Social Problem Solving Inventory, the Social Skills Inventory, and a series of questions about the degree to which anger affected their lives. Symptomatology was measured by the Brief Symptom Inventory and served as the criterion measure for a series of multiple regression analyses. Results indicated that both anger and social factors related to measures of psychological distress. Implications of the relationship between anger, social skills, and psychological symptoms are discussed in terms of research and assessment of individuals who may suffer from anger problems.

Behavior, Attitudes, and Cognitions of Anger-Prone Individuals

This research explored facets of anger based on a multidimensional-associationistic conceptualization (Berkowitz, 1994) that includes antecedents, behavior, cognitions, and experiential response dimensions. High and low anger-prone individuals responded to six audiotaped situations validated in previous work to be anger provoking. Participants evaluated their own audiotaped responses, as did peer judges. Participants also completed the Social Problem Solving Inventory (SPSI) and the Brief Symptom Inventory (BSI) and they reported on the effect of anger experiences in their lives. Neither self nor peer ratings of the audiotaped responses differentiated the behavior of anger groups. High and low anger-prone individuals, however, differed in the way that they viewed the consequences of their behavior and how anger affected them in general. Furthermore, high anger-prone individuals had elevated levels of pathology on eight of the nine scales of the BSI and scored significantly lower on the SPSI. Results are discussed in terms of assessment of anger proneness, its relationship to psychopathology and implications for future work.

Social Anxiety and Social Performance: Why Don't We See More Catastrophes?

This article surveys the recent literature with regard to social phobia and social anxiety as they affect social performance. It suggests that some of the inconsistent findings in this area might be alleviated by the application of Fazey and Hardys (1988) catastrophe model to ongoing research in the area of social performance.