Estimating the causal effects of treatment

OBJECTIVE

To provide a relatively non-technical review of recent statistical research on the analysis and interpretation of the results of randomised controlled trials in which there are possibly all three of the following types of protocol violation: non-adherence to allocated treatment, contamination (that is, patients receiving treatments other than the one to which they were allocated) and attrition (missing outcome data).

METHODS

The estimation methods involve the use of potential outcomes (counterfactuals) in the definition of a causal effect of treatment and in drawing valid inferences concerning its size.

RESULTS

The methods are explained through the use of simple arithmetical expressions involving the counts from three-way contingency tables (Outcome by Treatment Received by Random Allocation). Illustration is provided through the use of a hypothetical data set.

CONCLUSIONS

Recent advances in statistical methodology enable one to estimate treatment effects from the results of randomized trials in which the treatment actually received is not necessarily the one to which the patient was allocated. These methods allow one to make adjustments to allow for both non-compliance and loss to follow-up. Even for such a 'broken' randomised trial, inference concerning causal effects is safer than that from data arising from an observational study that never involved random allocation in the first place.

Acute effects of oral cocaine on inhibitory control of behavior in humans

Studies of humans show that individuals with histories of cocaine abuse display reduced inhibitory control over behavioral impulses. The present study tested the effects of oral cocaine on the ability to inhibit behavior in humans. Eight adult volunteers (seven men and one woman) with a history of cocaine abuse participated as in-patient volunteers. Response inhibition and response execution were measured by a stop-signal paradigm using a choice reaction time task that engaged subjects in responding to go-signals when stop-signals occasionally informed them to inhibit the response. Subjects' performance on the task was tested just before and 1 h after a randomized, double-blind administration of 0 mg (placebo), 50, 100, and 150 mg of oral cocaine HCl. Cocaine reduced subjects' ability to inhibit responses to stop-signals. By contrast, no effect of cocaine was observed on the ability to execute responses in terms of their speed and accuracy. Subjective and physiological effects of cocaine were also observed. Together, the findings indicate that acute administration of cocaine can impair the ability to inhibit behavioral responses at doses that do not affect the ability to respond. These findings are important because they identify a specific disinhibiting effect of cocaine that could help explain the documented association between long-term cocaine use and poor impulse control.

Abstinence reinforcement maintenance contingency and one-year follow-up

BACKGROUND

Relapse to drug use is often seen when contingencies designed to reduce drug use are discontinued. This paper reports on a stepdown maintenance contingency and 1-year follow-up in 110 patients who were maintained on methadone (50 or 70 mg/day) and who had completed a contingency management trial targeted to decreasing their opiate use. In the prior study (induction phase, 8 weeks) participants received vouchers for each opiate-negative urine screen or noncontingently.

METHODS

In this study (maintenance phase, 12 weeks), participants were rerandomized to receive vouchers and take-home methadone doses contingent on providing opiate-negative urine specimens (N=55) or noncontingently (N=55). Since participants had been rerandomized from induction-phase contingencies, most study data were analyzed as if from a 2 x 2 (inductionxmaintenance) design. Follow-up interviews were conducted at 3, 6, and 12 months after study participation.

RESULTS

Patients who received the maintenance contingency following an 8-week induction contingency had better outcomes than those who received noncontingent incentives in either the maintenance or induction phases of the trial. Good outcome at follow-up was predicted by enrollment in methadone maintenance after the study. Significantly more participants in the maintenance contingency group transferred directly to another methadone program.

CONCLUSION

These findings support the therapeutic value of extending the duration of contingency management and long-term methadone maintenance.

Dynamics and constraints in insight problem solving

This article reports 2 experiments that investigated performance on a novel insight problem, the 8-coin problem. The authors hypothesized that participants would make certain initial moves (strategic moves) that seemed to make progress according to the problem instructions but that nonetheless would guarantee failure to solve the problem. Experiment 1 manipulated the starting state of the problem and showed that overall solution rates were lower when such strategic moves were available. Experiment 2 showed that failure to capitalize on visual hints about the correct first move was also associated with the availability of strategic moves. The results are interpreted in terms of an information-processing framework previously applied to the 9-dot problem. The authors argue that in addition to the operation of inappropriate constraints, a full account of insight problem solving must incorporate a dynamic that steers solution-seeking activity toward the constraints.

The development of impulsivity, fearfulness, and helpfulness during childhood: patterns of consistency and change in the trajectories of boys and girls

BACKGROUND

The objective of the present study was to describe the development of boys and girls during the elementary-school years on three dimensions that conceptually and empirically represent risk for maladjustment.

METHOD

Every year between kindergarten and grade six, teachers rated the impulsivity, fearfulness, and helpfulness dimensions among a sample of 1,865 children representative of kindergarten boys and girls in the province of Quebec (Canada) in 1986-87. A group-based trajectory method was used to 1) identify groups of boys and girls following distinct-level trajectories of behaviours (on each dimension) during the elementary-school years; 2) estimate the proportion of children in each of the identified trajectory groups; and 3) estimate the patterns of consistency and variations in trajectories.

RESULTS

The results indicated that the best models comprised three distinct-level trajectory groups on fearfulness and helpfulness (a low, moderate, and high group) and four distinct-level trajectory groups on impulsivity. The helpfulness and fearfulness trajectory groups were generally more stable than the impulsivity groups. The broad patterns of development were similar across sexes. However, there were more boys on the higher impulsivity trajectories and low helpfulness trajectory, while there were more girls on the high fearfulness trajectory.

CONCLUSION

We found that behavioural consistency over middle childhood varied across trajectory groups and across dimensions, and we identified sex differences in the distribution of children in the different trajectory groups that may reflect gender-specific risks for psychopathology.

Assessing dissociation as a risk factor for posttraumatic stress disorder: a study of adult offspring of holocaust survivors

Dissociative symptoms are frequently present in trauma survivors with posttraumatic stress disorder (PTSD). However, the possibility that dissociative symptoms may comprise a risk factor for the development of PTSD has not been examined. The current research investigates this possibility by evaluating dissociative symptoms in a group of adult offspring of Holocaust survivors, whom we have previously shown to be at increased risk of PTSD. Eighty-seven Holocaust survivor offspring and 39 comparison participants completed the Dissociative Experiences Scale, and assessments of trauma exposure, psychopathology, and parental PTSD. Dissociative symptoms were elevated in individuals with current PTSD, but not in those with past PTSD or with the risk factor of parental PTSD. Dissociative symptoms were also associated with forms of psychopathology other than PTSD. The results suggest that dissociative symptoms are related to current psychiatric symptomatology, including PTSD, rather than representing an enduring trait or preexisting risk factor for the development of PTSD.

Cross-sectionally and longitudinally balanced effects of processing speed on intellectual abilities

The processing speed theory of cognitive aging states that declines in intellectual abilities other than processing speed are mainly due to a slowing of processing speed. Cross-sectional studies have provided support for the processing speed theory as for age-related cognitive differences. Longitudinal studies, in turn, have provided weaker support for the processing speed theory as for cognitive age changes. The present study aims to reconcile this discrepancy of cross-sectional and longitudinal results by constructing a "fair" test of the processing speed theory, i.e., a test that balances the cross-sectional age range and the time period covered longitudinally. Data from 83 older adults came from the Bonn Longitudinal Study on Aging. Using hierarchical linear models, it is shown that, although cross-sectional age differences and longitudinal changes in the intellectual abilities were equal, processing speed only attenuated cross-sectional age effects in cognition but virtually did not reduce longitudinal change effects in cognition. This persisting difference in the explanatory power of processing speed regarding age-related differences and age-related changes is discussed with reference to other longitudinal studies and statistical issues regarding cross-sectional and longitudinal studies of aging.

Applying the Bollen-Stine Bootstrap for Goodness-of-Fit Measures to Structural Equation Models with Missing Data

The study proposed a method for extending the Bollen-Stine bootstrap of model fit to structural equation models with missing data. Matrix algebra difficulties associated with an incomplete data matrix are circumvented by applying the Bollen-Stine transformation to each case (or group of cases sharing a common pattern of missing data) using reduced arrays that contain elements corresponding to the observed variables. A SAS macro program is provided for the purposes of implementing this procedure, and its' performance was assessed in a simulation that varied distribution shape, sample size, and the missing data rate. Compared to the unadjusted fit statistic, which produced dramatically inflated Type I error rates, the bootstrap yielded model rejection rates quite close to the nominal 5% level, although rejection rates were conservative under small sample conditions.

Testing stage effects in an ethnically diverse sample

The Transtheoretical Model (TTM) has been extensively validated in representative samples of adult smokers. Stage effects, i.e., the patterned relationships between Stage of Change (SOC) and other TTM variables, have been reported in a variety of samples. This study describes reliability data for the TTM variables and tests the stage effects with an ethnically diverse sample of 296 parents with young children. On the basis of theory and previous empirical evidence from a general population, it is possible to make quantitative predications about the magnitude of the expected effect sizes for the Decisional Balance, Situational Temptations, and Processes of Change subscales. For each variable, both a test of significance and a comparison with the expected effect sizes is reported. Results indicated significant stage differences for 8 of the 11 TTM subscales, and all quantitative predictions were confirmed. This study supports the use of TTM measures in an ethnically diverse sample.

Theory development should begin (but not end) with good empirical fits: a comment on Roberts and Pashler (2000)

S. Roberts and H. Pashler (2000) argued against using goodness of fit as evidence to support theories. The authors agree with their suggestions for how to go beyond good fits but disagree with their starting point. In this comment, the authors argue that good fits are part and parcel of theory development, that they are part and parcel of the processes suggested by S. Roberts and H. Pashler, and that they must be the starting point (though far from the ending point) in theoretical development. The authors discuss historical examples of scientific theory development, recent examples of psychological theory development, and development of a particular theory (social contagion theory; J. L. Rodgers & D. C. Rowe, 1993) that S. Roberts and H. Pashler criticized.

Measuring change: a commentary on Keith et al. (2002)

A study by J. R. Keith et al. (2002) purported to find postoperative changes in attention in a sample of patients who underwent cardiopulmonary bypass surgery. Detecting differential patterns of neuropsychological change following treatment presents formidable methodological and statistical challenges. This commentary examines the findings of J. R. Keith et al. with regard to (a) the feasibility of randomizing research participants into treatment and control groups, (b) the handling of missing data, (c) the selection of statistical techniques to measure change, and (d) the measurement of differential neuropsychological deficit. Alternative strategies are discussed.

Taxonicity of adolescent melancholia: a categorical or dimensional construct?

A taxometric analysis was conducted to test the hypothesis that the latent structure of melancholia in adolescents is categorical. Two taxometric procedures were used: Mean Above Minus Below a Cut (MAMBAC) and Maximum Covariance (MAXCOV) analyses. Participants were 378 adolescents presenting for a depression evaluation. Indicators of melancholia were constructed using items from the Schedule for Affective Disorders and Schizophrenia for School Aged Children (K-SADS) and the Beck Depression Inventory (BDI). The indicators of melancholia were consistent with a categorical latent variable. The findings suggest that the latent structure of melancholia in adolescents is similar to its previously identified categorical structure in adults. Implications for clinical research are discussed.

Überblicksarbeit

Zusammenfassung. Präventionen für Hoch-Risiko-Gruppen (“targeted interventions“) erweisen sich nur dann als wirksam, wenn mit ihrer Hilfe die anvisierten Risikobedingungen in einer Population angesprochen werden. Ein Erfolg hängt somit von der Genauigkeit und Stärke der Methoden zur Identifikation der Hoch-Risiko-Kinder ab. Der effektive Einsatz von Präventionen setzt unter anderem voraus, dass ein Verständnis dafür besteht, welche Risikobedingungen zu welchem Entwicklungszeitpunkt vorliegen und wie stark die Beziehung zwischen Risikobedingung und einem Outcome (z.B. aggressives Verhalten) ist. Im vorliegenden Review wird zunächst ein Überblick zu den etablierten Entwicklungsmodellen und bekannten Risikobedingungen aggressiv/dissozialen Verhaltens gegeben und diese vor dem Hintergrund aktueller empirischer Befunde diskutiert. Metaanalytische Befunde und unterschiedliche methodische Ansätze verdeutlichen, dass für die Vorhersage eines späteren aggressiv/dissozialen Verhaltens nicht nur Verhaltensmerkmale (z.B. frühes externalisierendes Verhalten), sondern auch psychosoziale Risikobedingungen bedeutsam sind. Abschließend wird der Nutzen dieser Befunde für die Vorhersage eines aggressiv/dissozialen Verhaltens und die Durchführung von Multiple-Gating-Screeningprozeduren zur Identifikation von Hoch-Risiko-Kindern erörtert und das Problem von Fehlklassifizierungen in Vorhersagen vor dem Hintergrund von Befunden aus der Entwicklungspsychopathologie diskutiert.

Aberrant driving behaviour: homogeneity of a four-factor structure in samples differing in age and gender

Four samples, representing a broad age range, were compared in confirmatory factor analyses with respect to a four-factor model of aberrant driving behaviour. With the restrictions imposed on the data by the model, an approximate fit was obtained for all four samples. Additional analyses, which tested equality of factor loadings and correlations for different age groups, also indicated that the model was a good fit. Separate analyses on male and female respondents yielded similar results. Although earlier results have shown that aberrant driving behaviours are differentially related to age and gender, the findings suggest that structural differences in relation to age and gender may be of minor importance. The four-factor model of aberrant driving behaviour sufficiently meets statistical criteria in a model-generating phase. In addition, it is an appropriate solution to apply on data obtained from respondents varying in age and gender.

Response styles in affect ratings: making a mountain out of a molehill

Ratings of affect words are the most commonly used method to assess pleasant affect (PA) and unpleasant affect (UA). The reliance on self-reports would be problematic if affect ratings were heavily influenced by response styles. Several recent publications have indeed suggested (a) that the influence of response styles on affect ratings is pervasive, (b) that this influence can be controlled by variations of the response format using multitrait-multimethod models, and (c) the discriminant validity of PA and UA is spurious. In this article, we examined the evidence for these claims. We demonstrate that (a) response styles have a negligible effect on affect ratings, (b) multiple response formats produce the same results as a single response format, and (c) the discriminant validity of PA and UA is not a method artifact. Rather, evidence against discriminant validity is due to the use of inappropriate response formats that respondents interpreted as bipolar scales.

A comparison of maximum covariance and K-means cluster analysis in classifying cases into known taxon groups

Maximum covariance (MAXCOV) is a method for determining whether a group of 3 or more indicators marks 1 continuous or 2 discrete latent distributions of individuals. Although the circumstances under which MAXCOV is effective in detecting latent taxa have been specified, its efficiency in classifying cases into groups has not been assessed, and few studies have compared its performance with that of cluster analysis. In the present Monte Carlo study, the classification efficiencies of MAXCOV and the k-means algorithm were compared across ranges of sample size, effect size, indicator number, taxon base rate, and within-groups covariance. When the impact of these parameters was minimized, k-means classified more data points correctly than MAXCOV. However, when the effects of all parameters were increased concurrently, MAXCOV outperformed k-means.

Dialectical behavior therapy versus comprehensive validation therapy plus 12-step for the treatment of opioid dependent women meeting criteria for borderline personality disorder

We conducted a randomized controlled trial to evaluate whether dialectical behavior therapy (DBT), a treatment that synthesizes behavioral change with radical acceptance strategies, would be more effective for heroin-dependent women with borderline personality disorder (N = 23) than Comprehensive Validation Therapy with 12-Step (CVT + 12S), a manualized approach that provided the major acceptance-based strategies used in DBT in combination with participation in 12-Step programs. In addition to psychosocial treatment, subjects also received concurrent opiate agonist therapy with adequate doses of LAAM (thrice weekly; modal dose 90/90/130 mg). Treatment lasted for 12 months. Drug use outcomes were measured via thrice-weekly urinalyses and self-report. Three major findings emerged. First, results of urinalyses indicated that both treatment conditions were effective in reducing opiate use relative to baseline. At 16 months post-randomization (4 months post-treatment), all participants had a low proportion of opiate-positive urinalyses (27% in DBT; 33% in CVT + 12S). With regard to between-condition differences, participants assigned to DBT maintained reductions in mean opiate use through 12 months of active treatment while those assigned to CVT + 12S significantly increased opiate use during the last 4 months of treatment. Second, CVT + 12S retained all 12 participants for the entire year of treatment, compared to a 64% retention rate in DBT. Third, at both post-treatment and at the 16-month follow-up assessment, subjects in both treatment conditions showed significant overall reductions in level of psychopathology relative to baseline. A noteworthy secondary finding was that DBT participants were significantly more accurate in their self-report of opiate use than were those assigned to CVT + 12S.

Commitment to organizational change: extension of a three-component model

Three studies were conducted to test the application of a three-component model of workplace commitment (J. P. Meyer & N. J. Allen, 1991: J. P. Meyer & L. Herscovitch, 2001) in the context of employee commitment to organizational change. Study 1, conducted with 224 university students, provided preliminary evidence for the validity of newly developed Affective, Continuance, and Normative Commitment to Change Scales. Studies 2 and 3, conducted with hospital nurses (N = 157 and 108, respectively), provided further support for the validity of the three Commitment to Change Scales, and demonstrated that (a) commitment to a change is a better predictor of behavioral support for a change than is organizational commitment, (b) affective and normative commitment to a change are associated with higher levels of support than is continuance commitment, and (c) the components of commitment combine to predict behavior.

Measurement equivalence: a comparison of methods based on confirmatory factor analysis and item response theory

Current interest in the assessment of measurement equivalence emphasizes 2 major methods of analysis. The authors offer a comparison of a linear method (confirmatory factor analysis) and a nonlinear method (differential item and test functioning using item response theory) with an emphasis on their methodological similarities and differences. The 2 approaches test for the equality of true scores (or expected raw scores) across 2 populations when the latent (or factor) score is held constant. Both approaches can provide information about when measurement nonequivalence exists and the extent to which it is a problem. An empirical example is used to illustrate the 2 approaches.

Deficient response modulation and emotion processing in low-anxious Caucasian psychopathic offenders: results from a lexical decision task

The clinical and research literatures on psychopathy have identified an emotion paradox: Psychopaths display normal appraisal but impaired use of emotion cues. Using R. D. Hare's (1991) Psychopathy Checklist-Revised and the G. S. Welsh Anxiety Scale (1956), the authors identified low-anxious psychopaths and controls and examined predictions concerning their performance on a lexical-decision task. Results supported all the predictions: (a) low-anxious psychopaths appraised emotion cues as well as controls; (b) their lexical decisions were relatively unaffected by emotion cues; (c) their lexical decisions were relatively unaffected by affectively neutral word-frequency cues; and (d) their performance deficits were specific to conditions involving right-handed responses. The authors propose that deficient response modulation may underlie both the emotional and cognitive deficits associated with low-anxious psychopaths.

Substance user treatment program quality: selected topics

This panel explores the "state of the art" in conceptualization and research pertinent to program quality in substance user treatment. First, seven critical questions for program quality are identified and discussed. Second, a recent national evaluation of treatment examines the implications of long-term patient outcomes for treatment quality. Third, a large number of clinical trials of behavioral and pharmacological treatments for cocaine dependence were conducted during the 1990s; this research is synthesized and interpreted. Fourth, progress is reported in improving the quality of treatment through standardized criteria for patient placement matching.

Development of the practice environment scale of the Nursing Work Index

Five subscales were derived from the Nursing Work Index (NWI) to measure the hospital nursing practice environment, using 1985-1986 nurse data from 16 magnet hospitals. The NWI comprises organizational characteristics of the original magnet hospitals. The psychometric properties of the subscales and a composite measure were established. All measures were highly reliable at the nurse and hospital levels. Construct validity was supported by higher scores of nurses in magnet versus nonmagnet hospitals. Confirmatory analyses of contemporary data from 11,636 Pennsylvania nurses supported the subscales. The soundness of the new measures is supported by their theoretical and empirical foundations, conceptual integrity, psychometric strength, and generalizability. The measures could be used to study how the practice environment influences nurse and patient outcomes.

Five-factor model of personality and job satisfaction: a meta-analysis

This study reports results of a meta-analysis linking traits from the 5-factor model of personality to overall job satisfaction. Using the model as an organizing framework, 334 correlations from 163 independent samples were classified according to the model. The estimated true score correlations with job satisfaction were -.29 for Neuroticism, .25 for Extraversion, .02 for Openness to Experience, .17 for Agreeableness, and .26 for Conscientiousness. Results further indicated that only the relations of Neuroticism and Extraversion with job satisfaction generalized across studies. As a set, the Big Five traits had a multiple correlation of .41 with job satisfaction, indicating support for the validity of the dispositional source of job satisfaction when traits are organized according to the 5-factor model.

Analysis of variance frameworks in clinical child and adolescent psychology: advanced issues and recommendations

Explores more advanced issues that researchers need to consider when using analysis of variance frameworks, building on basic issues for analysis of variance discussed in Jaccard and Guilamo-Ramos (2002). These include (a) using confidence intervals, (b) asserting group equivalence after a nonsignificant result, (c) use of magnitude estimation approaches, (d) sample size and power considerations, (e) outlier analysis, (f) violations of assumptions, and (g) missing data. Suggestions are offered for analytic practices in each of these domains.

On the bias of Huffcutt and Arthur's (1995) procedure for identifying outliers in the meta-analysis of correlations

This study documents how the use of A. I. Huffcutt & W. A. Arthur's (1995) sample adjusted meta-analytic deviancy (SAMD) statistic for identifying outliers in correlational meta-analyses results in inaccuracies in mean r. Monte Carlo simulations found that use of the SAMD resulted in the overidentification of small relative to large correlations as outliers. Furthermore, this tendency to overidentify small correlations was found to increase as the magnitude of the population correlation increased and resulted in mean rs that overestimated the population correlation. The implications for meta-analysts are discussed, and 2 possible solutions are offered.

A critique of the correlated trait-correlated method and correlated uniqueness models for multitrait-multimethod data

The correlated trait-correlated method (CT-CM) and correlated uniqueness (CU) confirmatory factor analysis models for multitrait-multimethod data are critiqued. Although the CU model often returns convergent and admissible factor solutions when the CT-CM model does not, the CU model is shown to have theoretical and substantive shortcomings. On the basis of this critique, the authors recommend that the CT-CM model be regarded as the generally preferred model and that the CU model be invoked only when the CT-CM model fails.

Processing of gene expression data generated by quantitative real-time RT-PCR

Quantitative real-time PCR represents a highly sensitive and powerful technique for the quantitation of nucleic acids. It has a tremendous potential for the high-throughput analysis of gene expression in research and routine diagnostics. However, the major hurdle is not the practical performance of the experiments themselves but rather the efficient evaluation and the mathematical and statistical analysis of the enormous amount of data gained by this technology, as these functions are not included in the software provided by the manufacturers of the detection systems. In this work, we focus on the mathematical evaluation and analysis of the data generated by quantitative real-time PCR, the calculation of the final results, the propagation of experimental variation of the measured values to the final results, and the statistical analysis. We developed a Microsoft Excel-based software application coded in Visual Basic for Applications, called Q-Gene, which addresses these points. Q-Gene manages and expedites the planning, performance, and evaluation of quantitative real-time PCR experiments, as well as the mathematical and statistical analysis, storage, and graphical presentation of the data. The Q-Gene software application is a tool to cope with complex quantitative real-time PCR experiments at a high-throughput scale and considerably expedites and rationalizes the experimental setup, data analysis, and data management while ensuring highest reproducibility.

Power in randomized group comparisons: the value of adding a single intermediate time point to a traditional pretest-posttest design

Adding a pretest as a covariate to a randomized posttest-only design increases statistical power, as does the addition of intermediate time points to a randomized pretest-posttest design. Although typically 5 waves of data are required in this instance to produce meaningful gains in power, a 3-wave intensive design allows the evaluation of the straight-line growth model and may reduce the effect of missing data. The authors identify the statistically most powerful method of data analysis in the 3-wave intensive design. If straight-line growth is assumed, the pretest-posttest slope must assume fairly extreme values for the intermediate time point to increase power beyond the standard analysis of covariance on the posttest with the pretest as covariate, ignoring the intermediate time point.

When constraints interact: a caution about reference variables, identification constraints, and scale dependencies in structural equation modeling

In traditional approaches to structural equations modeling, variances of latent endogenous variables cannot be specified or constrained directly and, consequently, are not identified, unless certain precautions are taken. The usual method for achieving identification has been to fix one factor loading for each endogenous latent variable at unity. An alternative approach is to fix variances using newer constrained estimation algorithms. This article examines the philosophy behind such constraints and shows how their appropriate use is neither as straightforward nor as noncontroversial as portrayed in textbooks and computer manuals. The constraints on latent variable variances can interact with other model constraints to interfere with the testing of certain kinds of hypotheses and can yield incorrect standardized solutions with some popular software.

Two studies examining the negative effect of self-efficacy on performance

Although hundreds of studies have found a positive relationship between self-efficacy and performance, several studies have found a negative relationship when the analysis is done across time (repeated measures) rather than across individuals. W. T. Powers (1991) predicted this negative relationship based on perceptual control theory. Here, 2 studies are presented to (a) confirm the causal role of self-efficacy and (b) substantiate the explanation. In Study 1, self-efficacy was manipulated for 43 of 87 undergraduates on an analytic game. The manipulation was negatively related to performance on the next trial. In Study 2, 104 undergraduates played the analytic game and reported self-efficacy between each game and confidence in the degree to which they had assessed previous feedback. As expected, self-efficacy led to overconfidence and hence increased the likelihood of committing logic errors during the game.

Statistical power in randomized intervention studies with noncompliance

This study examined various factors that affect statistical power in randomized intervention studies with noncompliance. On the basis of Monte Carlo simulations, this study demonstrates how statistical power changes depending on compliance rate, study design, outcome distributions, and covariate information. It also examines how these factors influence power in different methods of estimating intervention effects. Intent-to-treat analysis and complier average causal effect estimation are compared as 2 alternative ways of estimating intervention effects under noncompliance. The results of this investigation provide practical implications in designing and evaluating intervention studies taking into account noncompliance.

Measurement of functional ability following traumatic brain injury using the Clinical Outcomes Variable Scale: a reliability study

This study determined the inter-tester and intra-tester reliability of physiotherapists measuring functional motor ability of traumatic brain injury clients using the Clinical Outcomes Variable Scale (COVS). To test inter-tester reliability, 14 physiotherapists scored the ability of 16 videotaped patients to execute the items that comprise the COVS. Intra-tester reliability was determined by four physiotherapists repeating their assessments after one week, and three months later. The intra-class correlation coefficients (ICC) were very high for both inter-tester reliability (ICC > 0.97 for total COVS scores, ICC > 0.93 for individual COVS items) and intra-tester reliability (ICC > 0 97). This study demonstrates that physiotherapists are reliable in the administration of the COVS.

Jointly modelling the relationship between survival and pulmonary function in cystic fibrosis patients

Modelling the relationship between pulmonary function and survival in cystic fibrosis (CF) is complicated by the fact that measures of pulmonary function commonly used such as the forced expiratory volume in one second (FEV(1)) are measured with error, and patients with the poorest lung function are increasingly censored by death, that is, data are available only for the patients who have survived to the current age. We assume a linear random effects model for FEV1 per cent predicted, where the random intercept and slope of FEV(1) per cent predicted, along with a specified transformation of the age at death follow a trivariate normal distribution. We illustrate how this model can be used to describe the relationship between age at death and parameters of the individual patient's regression of FEV(1) per cent predicted versus age, such as the slope and the intercept or true value of FEV(1) per cent predicted at a given age. We also illustrate how the model provides empirical Bayes estimates of these individual parameters. In particular, we explore how the predicted value of the age at death might be used as a prognostic or severity index. The model and methods are illustrated on a cohort of 188 cystic fibrosis patients with a common genotype (homozygous for the DeltaF508 mutation), born on or after 1965 and followed at the CF Center at the Rainbow Babies and Children's Hospital, Cleveland, OH, U.S.A.

Sample size requirements for estimating intraclass correlations with desired precision

A method is developed to calculate the approximate number of subjects required to obtain an exact confidence interval of desired width for certain types of intraclass correlations in one-way and two-way ANOVA models. The sample size approximation is shown to be very accurate.

Power dressing and meta-analysis: incorporating power analysis into meta-analysis

AIMS

This paper highlights the lack of consideration that is given to power in the health and social sciences, which is a continuing problem with both single study research and more importantly for meta-analysis.

BACKGROUND

The power of a study is the probability that it will lead to a statistically significant result. By ignoring power the single study researcher makes it difficult to get negative results published and therefore affects meta-analysis through publication bias. Researchers using meta-analysis, who also ignore power, then compound the problem by including studies with low power that are more likely to show significant effects.

METHOD

A simple means of calculating an easily understood measure of effect size from a contingency table is demonstrated in this paper. A computer programme for determining the power of a study is recommended and a method of reflecting the adequacy of the power of the studies in a meta-analysis is suggested. An example of this calculation from a meta-analytic study on intravenous magnesium, which produced inaccurate results, is provided.

CONCLUSION

It is demonstrated that incorporating power analysis into this meta-analysis would have prevented misleading conclusions being reached. Some suggestions are made for changes in the protocol of meta-analytic studies, which highlight the importance of power analysis.

Effect of harmonic relatedness on the detection of temporal asynchronies

Speeded intonation judgments of a target chord are facilitated when the chord is preceded by a harmonically related prime chord. The present study extends harmonic priming to temporal asynchrony judgments. In both tasks, the normative target chords (consonant, synchronous) are processed more quickly and accurately after a harmonically related prime than after a harmonically unrelated prime. However, the influence of harmonic context on sensitivity (d') differs between the two tasks: d' was higher in the related context for intonation judgments but was higher in the unrelated context for asynchrony judgments. A neural net model of tonal knowledge activation provides an explanatory framework for both the facilitation in the related contexts and the sensitivity differences between the tasks.

Abnormal P300 in a case of delusional misidentification with coinciding Capgras and Frégoli symptoms

The Delusional Misidentification Syndrome (DMS) is thought to be related to dissociation between recognition and identification processes. Working memory (WM) is considered responsible for the integration and online manipulation of information, so that it is available for further processing. Since the P300 component of event-related potentials (ERPs) is considered as an index of the on-line updating of WM, the present study is focused on auditory P300 elicited during a WM test in DMS, compared with that in healthy controls. ERPs, elicited during a WM test, in a case suffering from coinciding Capgras and Frégoli symptoms, were recorded. Peak amplitude and latency of the averaged P300 waveforms, as well as memory performance of this case, were compared to the patterns obtained from healthy controls. In relation to normal controls, the patient exhibited significantly attenuated amplitude of P300 at the F4, P3 and Pz abductions. The patient also showed significantly prolonged latencies of P300 at all abductions used. These findings suggest that DMS may be accompanied by WM dysfunction affecting brain regions outside the prefrontal cortex, as well as within, and by diffuse failure to allocate attention resources to a stimulus, as they are reflected by P300 amplitudes and latencies respectively. Additionally, it may be suggested that techniques designed to explore cognitive operations, such as recording of ERPs, and more specifically P300, during WM tasks, could provide further insights into the relationship between neural functioning and the cognitive deficits in DMS.