A general model for the analysis of multilevel data

A general model is developed for the analysis of multivariate multilevel data structures. Special cases of the model include repeated measures designs, multiple matrix samples, multilevel latent variable models, multiple time series, and variance and covariance component models.We would like to acknowledge the helpful comments of Ruth Silver. We also wish to thank the referees for helping to clarify the paper. This work was partly carried out with research funds provided by the Economic and Social Research Council (U.K.).     

The importance of assessing treatment integrity: An example in the anxiety disorders

One purpose of this investigation was to examine the importance of assessing treatment integrity in a study evaluating a treatment for specific anxiety disorders. Three subjects, two social phobics and one simple phobic, received self-instruction training (Phase I) followed by exposure and self-instruction training (Phase II) in a multiple baseline across subjects design. All subjects were assessed during a pretreatment baseline and throughout treatment using measures of treatment integrity and measures of change in phobic severity. Improvement took place at different times for different subjects. Measures of treatment integrity, which consisted of monitoring of self-statements as well as practice outside the session, indicated that treatment was not always received as intended and that improvement was correlated with practice. The usefulness of measures of treatment integrity in both clinical and research settings is highlighted.     

Evaluating children's mental health service systems: An analysis of critical behaviors and events

A lack of studies which evaluate system change by tracking client outcomes is noted in the children's mental health area. This deficit may be a result of the inability of researchers to define outcomes and to draw conclusions about which measures reflect the efficacy of services and service delivery systems. This paper reviews five social validation surveys which examined children's mental health outcome measures. Based on the results of the five surveys, a model of critical behaviors and events is presented. This model will assist evaluators and researchers in understanding which critical events and behaviors should be measured to assess the impact of community-based mental health services for children.     

Factor analysis of the social performance survey schedule

There are several studies suggesting that the Social Performance Survey Schedule (SPSS) is a reliable and valid measure of social skill. However, the factorial structure of the SPSS has never been examined. In the present study, SPSSs taken by 652 college students were factor analyzed by the method of principal components and the resulting factors were rotated according to the varimax criterion. Separate factor analyses were conducted for men and women. Seven factors, which accounted for a little over a third of the total variance, were retained for each sex. A large general factor labeled Prosocial Competence emerged for both sexes. Beyond this, the factorial solutions for the two sexes tended to diverge, with males showing a large negative factor (Social Offensiveness) and three smaller negative factors and females showing five relatively small negative factors. Suggestions were made concerning additional factor analytic studies of the SPSS and the ways in which derived factors might be used in research and clinical work.This research was supported by funds from the Rutgers University Research Council. Thanks are due to Leona Aiken for her statistical consultation.     

APPLICATION OF A SOCIAL SKILLS TRAINING PROGRAM IN THE MODIFICATION OF INTERPERSONAL DEFICITS AMONG RETARDED ADULTS: A CLINICAL REPLICATION

Six mentally retarded adults, equally divided into two treatment groups, were provided with individualized social skills training programs. Treatment, evaluated via a multiple-baseline design strategy, was sequentially and cumulatively applied across target behaviors over a four-week intervention period. Behavioral observation probes and social validation measures served as the primary dependent variables. Results indicated that (a) treatment was effective for virtually all behaviors across all subjects, (b) improvements occurred for both training and generalization scenes, and (c) behavioral performance was maintained one month following the termination of treatment.     

ANALYSIS OF THE CONSISTENCY OF OBJECTIVE MEASURES OF SEXUAL AROUSAL IN WOMEN

Eight adult women volunteers viewed the same erotic film in two different sessions. Their resulting genital responses were recorded simultaneously by three different measures: vaginal pressure pulse, vaginal blood volume, and labial temperature change. During the erotic stimulation, the vaginal pressure pulse and labial responses always increased, and the vaginal blood volume response increased on all but two occasions. Although there was considerable intersubject variability in each genital measure, all three measures were found to have some intrasubject consistency over sessions with respect to either their response amplitudes or patterns, with labial temperature being the most consistent on both parameters. The relationship between the response patterns of the three measures during the film was also relatively consistent across sessions, as was the correspondence between subjective ratings of arousal and both vaginal pressure pulse and labial responses but not vaginal blood volume response. To overcome the problem of considerable intrasubject variability of response amplitudes, it was suggested that the inclusion in the data analysis of several parameters of response patterns, which were relatively stable over sessions, might facilitate the evaluation of a treatment.     

Considerations in the choice of interobserver reliability estimates

Two types of interobserver reliability values may be needed in treatment studies in which observers constitute the primary data-acquisition system: trial reliability and the reliability of the composite unit or score which is subsequently analyzed, e.g., daily or weekly session totals. Two approaches to determining interobserver reliability are described: percentage agreement and "correlational" measures of reliability. The interpretation of these estimates, factors affecting their magnitude, and the advantages and limitations of each approach are presented.     

Resting autonomic activity in adults who stutter and its association with self-reports of social anxiety

PurposeThe purpose of this study was to investigate resting autonomic activity in adults who stutter (AWS) compared to adults who do not stutter (ANS) and the relationship this has on self-reports of social anxiety.MethodsThirteen AWS and 15 ANS completed the Social Interaction Anxiety Scale (SIAS; Mattick & Clark, 1998) and Brief Fear of Negative Evaluation (BFNE; Leary, 1983). Following this, measures of skin conductance levels (i.e. index of sympathetic activity) and respiratory sinus arrhythmia (i.e. index of parasympathetic activity) were taken during a 5-minute resting, baseline period. Independent sample t tests were used to assess differences between groups on self-reports of anxiety (SIAS, BFNE) and resting autonomic levels (SCL, RSA). Separate multiple regression analyses were performed in order to assess the relationship between self-reports of anxiety and autonomic measures.ResultsResults showed significantly higher mean SCL and lower mean RSA levels in the AWS compared to the ANS at resting, baseline. Regression analysis showed that self-reports from the SIAS had a significant effect on RSA levels for the AWS but not the ANS. No significant effects were found for BFNE on RSA. Nor was there a significant effect from SIAS or BFNE on SCL levels for either group.ConclusionFindings suggest that resting RSA levels may be a physiological marker for social anxiety levels in adults who stutter.     

Caveats of non-ipsatization of basic values: A review of issues and a simulation study

Ipsatization, or a correction of variables by their common component, is routinely applied to measures of basic values. Although ipsatization has been criticized, the consequences of non-ipsatization are rarely discussed. We show that the ipsatization of values is intertwined with their definition. A common factor involved in ipsatization was suggested to represent a nuisance variable, a response style, social desirability, or other constructs. A simulation study illustrated that within-individual centering revealed more accurate value scores when the common factor was in the data, with exception of the situation when values were consistently and positively correlated with each other. We conclude that in different conditions both applying and failing to apply ipsatization can cause bias. Therefore, no general advice in regard to ipsatization can be provided.