Distorted correlations among censored data: causes, effects, and correction

Data censoring occurs when researchers do not know precise values of data points (e.g., age is 55+ or concentration ≤ .001). Censoring is frequent within psychology but typically unrecognized outside of longitudinal studies. We describe five circumstances when censoring may occur, demonstrate censoring distorts correlations, and discuss how censoring can create spurious factors. Next, we explain how to use R package lava to calculate maximum likelihood estimates (Holst and Budtz-Jørgensen Computational Statistics, 28(4), 1385-1452, 2013) of correlations between uncensored variables based upon censored variables. Previous research demonstrated these estimates were more accurate than Muthén's (1984) estimate for one particular model, but no research has systematically examined their accuracy. We therefore conducted a simulation study exploring the effects of the correlation, sample size, and censoring on point and interval estimates of correlations. Based upon 80 cells in which low values of normally distributed variables were censored, we recommend the constrained regression model with Wald confidence intervals. These methods were precise and unbiased unless both variables had 70% censoring and the correlation was large and negative (e.g., -.9), in which case estimates were closer to -1 than they should be. Opposite results would occur if low values of one variable and high values of the other were censored: Estimates would be precise and unbiased unless censoring was extreme and correlations were large and positive. To estimate large correlations accurately, we recommend researchers reduce censoring by using longer longitudinal studies, using scales with more response options, and matching measures to populations to reduce floor and ceiling effects.

Two and five-factor models of negative symptoms in schizophrenia are differentially associated with trait affect, defeatist performance beliefs, and psychosocial functioning

Recent factor analytic evidence supports both two-factor (motivation and pleasure, MAP; diminished expression, EXP) and five-factor (anhedonia, asociality, avolition, blunted affect, alogia) conceptualizations of negative symptoms. However, it is unclear whether these two conceptualizations of the latent structure of negative symptoms have differential associations with external correlates. The current study evaluated external correlates of the two- and five-factor structures by examining associations with variables known to have critical relations with negative symptoms: trait affect, defeatist performance beliefs, neurocognition, and community-based psychosocial functioning. Participants included a total of 245 outpatients diagnosed with schizophrenia who were rated on the Brief Negative Symptom Scale and completed a battery of additional measures during periods of clinical stability. These additional measures included the Positive and Negative Affect Schedule, Defeatist Performance Beliefs scale, MATRICS Consensus Cognitive Battery, and Level of Function Scale. Pearson correlations indicated differential patterns of associations between the BNSS scores and the external correlates. Support for the two-factor model was indicated by a stronger association of MAP with positive affect and psychosocial functioning, compared to EXP with neurocognition. Significance tests examining a differential magnitude of associations showed that the two-dimension negative symptom structure masked unique correlational relationships among the five negative symptom domains with neurocognition and social/vocational community functioning and captured unique patterns of correlation with trait affect. Support for the five-factor model was shown by a stronger association between Blunted Affect with Attention/Vigilance, and stronger associations between Avolition, Anhedonia, and Asociality with psychosocial functioning. Results offer support for both the two-dimension and five-domain model of negative symptoms as well as a hierarchical two-dimensions-five-domains model of negative symptoms. Findings may have implications for diagnostic criteria and descriptions of the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5), as well as possible treatment targets of negative symptoms.

Initial examination of the mental health disorders: screening instrument for athletes

Introduction

There is a need to psychometrically develop assessment instruments capable of screening mental health disorders in athlete populations. The current study was conducted to determine reliability, validity and clinical utility of the Mental Health Disorders Screening Instrument for Athletes (MHDSIA).

Methods and results

259 collegiate athletes completed the MHDSIA. Factor analysis determined a single factor with good internal consistency, and this factor was positively correlated with an established measure of psychiatric symptomology (Symptom Checklist 90-R), demonstrating its concurrent validity. An optimum clinical cutoff score (i.e., 32) was determined using Receiver Operating Characteristic (ROC) analyses to assist appropriate mental health referrals.

Discussion

Results suggest the MHSIA is a reliable, valid, and relatively quick and easy to interpret screen for the broad spectrum of mental health disorders in collegiate athletes. As expected, NCAA athletes reported lower MHDSIA scores than club and intramural athletes, while males reported similar severity scores as females.

Latent structure of cognitive tests is invariant in men and women with schizophrenia

Studies comparing the cognitive functioning of men and women with schizophrenia have produced conflicting results which could arise from sex-based differences in the latent structure of cognitive abilities. The current study used multigroup confirmatory factor analysis to examine invariance in latent structure of cognitive abilities to between men and women with schizophrenia. Confirmatory factor analysis of an initial neurocognitive assessment (men n = 612, women n = 201) and cross-validation using second assessment (men n = 549, women n = 198) demonstrated that a bifactor seven-factor model fit the data best for both men and women. Invariance analyses further indicated this model was invariant across men and women at both assessments. Group comparisons indicated women had significantly higher scores for Semantic Memory, Verbal Memory, and General Cognitive factors, whereas men exhibited better performance on the Vigilance factor. Results indicate that cognition in SZ is characterized by both a general cognitive factor and specific domains for both men and women. Invariance analysis provides evidence that cognitive differences between men and women do not result from sex-based differences in the latent structure of cognitive abilities. Current results also indicate small but statistically significant neurocognitive differences between men and women with schizophrenia.

Confirmatory factor analysis of imPACT cognitive tests in high school athletes

ImPACT is the most commonly utilized computerized neurocognitive assessment for the clinical management of sport concussion. The cognitive composite scores that ImPACT currently reports include Verbal Memory, Visual Memory, Visual Motor Speed, Reaction Time, and Impulse Control. However, exploratory factor analytic studies report that two or more factors may better represent ImPACT's latent structure, suggesting that the current cognitive composites may not adequately represent the cognitive constructs ImPACT assesses. The latent structure of ImPACT cognitive baseline scores was examined using exploratory (EFA) and confirmatory factor analysis (CFA) of valid baseline ImPACT scores for 36,091 high school athletes. These athletes were randomly divided into two samples. The first sample was a calibration sample used for EFA and the second sample was a cross-validation sample used for CFA to estimate the best model identified in the calibration phase, along with other models that were reported in the literature or based on theoretical considerations, including hierarchical and bifactor models. EFA identified a first-order four-factor solution consisting of Visual Memory, Visual Reaction Time, Verbal Memory, and Working Memory constructs. CFA indicated that this four-factor model provided superior fit for the data, while the current five-composite structure of ImPACT provided a poor fit for the data. The latent constructs identified in this study using CFA do not map well onto the composite scores that are currently used to interpret ImPACT performance. Future research should investigate whether interpretation of ImPACT based on the constructs identified here will be more useful for clinical decision making than current approaches. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Bifactor model of cognition in schizophrenia: Evidence for general and specific abilities

BACKGROUND

Despite extensive study of cognition in schizophrenia, it remains unclear as to whether cognitive deficits and their latent structure are best characterized as reflecting a generalized deficit, specific deficits, or some combination of general and specific constructs.

METHOD

To clarify latent structure of cognitive abilities, confirmatory factor analysis was used to examine the latent structure of cognitive data collected for the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) for Schizophrenia study. Baseline assessment data (n = 813) were randomly divided into calibration (n = 413) and cross-validation samples (n = 400). To examine whether generalized or specific deficit models provided better explanation of the data, we estimated first-order, hierarchical, and bifactor models.

RESULTS

A bifactor model with seven specific factors and one general factor provided the best fit to the data for both the calibration and cross-validation samples.

CONCLUSIONS

These findings lend support for a replicable bifactor model of cognition in schizophrenia, characterized by both a general cognitive factor and specific domains. This suggests that cognitive deficits in schizophrenia might be best understood by separate general and specific contributions.

Measuring the Perceived Social Intelligence of Robots

Robotic social intelligence is increasingly important. However, measures of human social intelligence omit basic skills, and robot-specific scales do not focus on social intelligence. We combined human robot interaction concepts of beliefs, desires, and intentions with psychology concepts of behaviors, cognitions, and emotions to create 20 Perceived Social Intelligence (PSI) Scales to comprehensively measure perceptions of robots with a wide range of embodiments and behaviors. Participants rated humanoid and non-humanoid robots interacting with people in five videos. Each scale had one factor and high internal consistency, indicating each measures a coherent construct. Scales capturing perceived social information processing skills (appearing to recognize, adapt to, and predict behaviors, cognitions, and emotions) and scales capturing perceived skills for identifying people (appearing to identify humans, individuals, and groups) correlated strongly with social competence and constituted the Mind and Behavior factors. Social presentation scales (appearing friendly, caring, helpful, trustworthy, and not rude, conceited, or hostile) relate more to Social Response to Robots Scales and Godspeed Indices, form a separate factor, and predict positive feelings about robots and wanting social interaction with them. For a comprehensive measure, researchers can use all PSI 20 scales for free. Alternatively, they can select the most relevant scales for their projects.

Expanding the cross-cultural psychological assessment tool box with IQ test short forms

Objective: Short forms of standard intelligence tests are useful in clinical and research settings where administration of a full intelligence test is impractical or unnecessary. In cross cultural contexts where few tests are available, including brief intelligence tests, short forms may be particularly useful to meet clinical and research needs. However, there is little cross-cultural research on the validity of short forms. This study evaluated feasibility of short form development for the Spanish version of the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) and examined comparability of short-form IQ estimates in Spanish- and English-speaking children with Attention-Deficit/Hyperactivity Disorder (ADHD).Method: Participants included children with ADHD who were administered the WISC-IV Spanish (n = 165) or WISC-IV (n = 299). Full Scale IQs (FSIQs) were calculated for two-, three-, four-, and five-subtest short forms. Misclassification rates, mean absolute differences, and intraclass correlation coefficients (ICCs) were used to determine short form accuracy in estimating FSIQ.Results: The WISC-IV short forms examined had generally strong psychometric properties (e.g., ICCs ranged from .78 to .94) and level of accuracy in estimating FSIQ did not differ across primary language of Spanish or English.Conclusions: Findings support feasibility of IQ short form development to help address mental health disparities in research and clinical screening for Spanish- and English-speaking pediatric populations with ADHD.

Factor Structure of the Wechsler Intelligence Scale for Children-Fourth Edition Spanish in a Clinical Sample

Objective

The Wechsler Intelligence Scale for Children-Fourth Edition Spanish (WISC-IV Spanish) is a translation and adaptation of the WISC-IV that is often used in neuropsychological evaluation of children and adolescents with acquired or developmental brain disorders. However, there is limited information on the factor structure of the WISC-IV Spanish when used with clinical populations. To address the potential effects of language and culture on the validity of the WISC-IV Spanish factor structure, this study examined the construct validity of the WISV-IV Spanish scores using confirmatory factor analysis in a clinical sample of Hispanic children with Attention-Deficit/Hyperactivity Disorder (ADHD).

Method

Participants included 148 children whose primary language was Spanish, who had a primary diagnosis of ADHD, and who had been administered the WISC-IV Spanish as part of a clinical evaluation (mean age = 10.6 years; SD = 2.7). Confirmatory factor analyses were performed to evaluate if the WISC-IV Spanish was best explained by a 1-, 2-, 3-, 4-, or 5-factor model based on the Cattell-Horn-Carroll theory.

Results

A 4-factor model composed of verbal comprehension, perceptual reasoning, working memory, and processing speed factors provided the best fit for the data (comparative fit index = .95; root mean square error of approximation = .06).

Conclusions

Findings for children with ADHD and LD support interpretation of WISC-IV Spanish Index scores based on the 4-factor model identified in Spanish- and English-speaking normative samples. Additional research with understudied clinical populations is warranted to address gaps in cross-cultural research.

Family contributions to sport performance and their utility in predicting appropriate referrals to mental health optimization programmes

Limited research has examined the contributions of sport-specific family relationship problems to athlete mental health. In the current study we examined the extent to which collegiate athletes' family problems (as measured by the Student Athlete Relationship Instrument factors, SARI) predict athletes' general mental health distress and specific mental health symptoms that are relatively common in athletes (i.e. depression, anxiety, and drug and alcohol use). The study included 85 collegiate athletes (intramural, n = 26; club sport, n = 12; NCAA Division I, n = 47). We hypothesized that both general mental health distress and commonly evidenced mental health symptoms would be predicted by athletes' responses to the SARI factors (Poor Relationship and Lack of Support, General Pressure, Pressure to Quit or Continue Unsafely, Embarrassing Comments, and Negative Attitude). Results indicated that all of the aforementioned SARI factors predicted athletes' ratings of depression and general mental health distress levels. Only the General Pressure SARI factor predicted athletes' anxiety and drug use. Alcohol use was not predicted by any of the SARI factors. Receiver operating characteristic (ROC) analyses indicated that sport-specific problems in family relationships provided a good classification of athletes at risk for general mental health distress, depression, and anxiety. Practice implications are discussed in light of the results.

The Latent Structure of Negative Symptoms in Schizophrenia

IMPORTANCE

Negative symptoms are associated with a range of poor clinical outcomes, and currently available treatments generally do not produce a clinically meaningful response. Limited treatment progress may be owing in part to poor clarity regarding latent structure. Prior studies have inferred latent structure using exploratory factor analysis, which has led to the conclusion that there are 2 dimensions reflecting motivation and pleasure (MAP) and diminished expressivity (EXP) factors. However, whether these conclusions are statistically justified remains unclear because exploratory factor analysis does not test latent structure. Confirmatory factor analysis (CFA) is needed to test competing models regarding the latent structure of a construct.

OBJECTIVE

To evaluate the fit of 4 models of the latent structure of negative symptoms in schizophrenia using CFA.

DESIGN, SETTING, AND PARTICIPANTS

Three cross-sectional studies were conducted on outpatients with schizophrenia who were rated on the 3 most conceptually contemporary measures: Scale for the Assessment of Negative Symptoms (SANS), Brief Negative Symptom Scale (BNSS), and Clinical Assessment Interview for Negative Symptoms (CAINS). Confirmatory factor analysis evaluated the following 4 models: (1) a 1-factor model; (2) a 2-factor model with EXP and MAP factors; (3) a 5-factor model with separate factors for the 5 domains of the National Institute of Mental Health consensus development conference (blunted affect, alogia, anhedonia, avolition, and asociality); and (4) a hierarchical model with 2 second-order factors reflecting EXP and MAP and 5 first-order factors reflecting the 5 consensus domains.

MAIN OUTCOMES AND MEASURES

Outcomes included CFA model fit statistics derived from symptom severity scores on the SANS, BNSS, and CAINS.

RESULTS

The study population included 860 outpatients with schizophrenia (68.0% male; mean [SD] age, 43.0 [11.4] years). Confirmatory factor analysis was conducted on each scale, including 268 patients for the SANS, 192 for the BNSS, and 400 for the CAINS. The 1- and 2-factor models provided poor fit for the SANS, BNSS, and CAINS as indicated by comparative fit indexes (CFIs) and Tucker Lewis indexes (TLIs) less than 0.950, RMSEAs that exceeded the 0.080 threshold, and WRMRs greater than 1.00. The 5-factor and hierarchical models provided excellent fit, with the 5-factor model being more parsimonious. The CFIs and TLIs met the 0.95 threshold and the 1.00 threshold for both factor models with all 3 measures. Interestingly, the RMSEAs for the 5-factor model and the hierarchical model fell under the 0.08 threshold for the BNSS and the CAINS but not the SANS.

CONCLUSIONS AND RELEVANCE

These findings suggest that the recent trend toward conceptualizing the latent structure of negative symptoms as 2 distinct dimensions does not adequately capture the complexity of the construct. The latent structure of negative symptoms is best conceptualized in relation to the 5 consensus domains. Implications for identifying pathophysiological mechanisms and targeted treatments are discussed.

Eating Disorder Quality of Life Scale (EDQLS) in ethnically diverse college women: an exploratory factor analysis

Background

Extant research suggests that disordered eating is common in college women and is associated with decreased quality of life. The Eating Disorder Quality of Life Scale (EDQLS) examines impairment to disordered eating-related quality of life, but has not been validated in college women. Accordingly, the purpose of this study was to examine the reliability, validity, and factor structure of the EDQLS in a diverse sample of 971 college women.

Method

Students from a large United States university completed questionnaires examining disordered eating and the EDQLS online.

Results

The EDQLS demonstrated excellent internal consistency and good convergent validity with the Eating Disorder Examination Questionnaire (EDEQ). Contrary to the original 12-domain design of the EDQLS, principal component analyses suggested five factors that mapped onto the following constructs: (1) Positive Emotionality; (2) Body/Weight Dissatisfaction; (3) Disordered Eating Behaviors; (4) Negative Emotionality; and (5) Social Engagement. However, 15 of the 40 items loaded onto multiple factors.

Conclusions

Total scores on the EDQLS are reliable and valid when used with diverse samples of college women, but some revisions are needed to create subscales than can justifiably be used in clinical practice.

Computer Scoring of Emotional Awareness in a Nonclinical Population of Young Adults

The Levels of Emotional Awareness Scale (LEAS; Lane, Quinlan, Schwartz, Walker, & Zeitlin, 1990 ) is an open-ended measure of the ability to describe emotional reactions. Scoring the LEAS by hand is complex and time consuming (Barchard, Bajgar, Leaf, & Lane, 2010 ). Therefore, Program for Open-Ended Scoring (POES; Leaf & Barchard, 2010 ) was designed to score the LEAS quickly and easily. Using 268 undergraduates, this article compares traditional LEAS hand scoring to 6 POES methods, 2 of which are holistic methods that have never before been examined. Based on split-half reliability, correlations with measures of emotional and social intelligence, and partial correlations once response length and vocabulary were partialed out, we recommend 3 of the POES methods when testing nonclinical samples of young adults. Because POES scoring is fast and efficient, it allows more researchers and clinicians to use the LEAS, thus moving away from self-report measures of emotional awareness.

Examination of the Extent to which Employment Factors are Associated with Reduced Child Maltreatment Potential and Drug Use

Job assistance programs are commonly recommended for parents of children who have been victimized by child maltreatment, particularly when illicit drugs are indicated. However, the relationship between employment factors, substance abuse and risk of child maltreatment has received limited empirical attention. Therefore, the current study examines employment factors in a sample of 72 mothers who were referred by Child Protective Services (CPS) for treatment of substance abuse and child neglect. Child maltreatment potential was found to be negatively associated with number of hours employed and self-reported happiness with employment. The association between child abuse potential and personal income of participants approached significance (p = .057), and the results were not influenced by social desirability. Employment satisfaction significantly contributed to the prediction of child maltreatment potential over and above other employment factors and control variables. These findings suggest that when mothers are involved in CPS their risk of perpetrating child maltreatment may be reduced when they are assisted in gainful employment that is personally satisfying. Happiness with employment was the only employment factor correlated (inversely) with substance use (biological testing, self-report of participants). Future directions are discussed in light of the results, including the importance of considering employment satisfaction when conducting vocational assistance programs in this population.

Equivalence of mother and father ratings of ADHD in children

Obtaining data from multiple informants provides a more comprehensive diagnostic picture in the assessment of attention deficit hyperactivity disorder (ADHD). Differences in symptom ratings have been observed between parent- and teacher-report scales, though less information is available regarding differences between mothers and fathers. To address this gap, this study examines the rater agreement between mothers and fathers on the Diagnostic and Statistical Manual of Mental Disorders - Fourth Edition (DSM-IV) ADHD Symptom Rating Scale (DSM-ADHD-SRS). The participants consisted of 337 children diagnosed with ADHD who underwent comprehensive neuropsychological assessment. Confirmatory factor analysis indicates that a three-factor model comprising inattention, hyperactivity, and impulsivity symptoms provides the best fit for both mothers' and fathers' ratings. Mothers provided higher mean ratings for the inattention scale. These results suggest that the factor structure for the DSM-ADHD-SRS is the same, regardless of parent gender. However, symptoms of inattention may vary depending upon which parent completes the ratings. This discrepancy could lead to differences in diagnostic impressions in clinical evaluations.

Internal Consistency and Power When Comparing Total Scores from Two Groups

Researchers now know that when theoretical reliability increases, power can increase, decrease, or stay the same. However, no analytic research has examined the relationship of power to the most commonly used type of reliability-internal consistency-and the most commonly used measures of internal consistency, coefficient alpha and ICC(A,k). We examine the relationship between the power of independent samples t tests and internal consistency. We explicate the mathematical model upon which researchers usually calculate internal consistency, one in which total scores are calculated as the sum of observed scores on K measures. Using this model, we derive a new formula for effect size to show that power and internal consistency are influenced by many of the same parameters but not always in the same direction. Changing an experiment in one way (e.g., lengthening the measure) is likely to influence multiple parameters simultaneously; thus, there are no simple relationships between such changes and internal consistency or power. If researchers revise measures to increase internal consistency, this might not increase power. To increase power, researchers should increase sample size, select measures that assess areas where group differences are largest, and use more powerful statistical procedures (e.g., ANCOVA).

Factor Structure of the Wechsler Intelligence Scale for Children: Fourth Edition in Children with ADHD

OBJECTIVE

Recent evidence suggests that the factor structure of the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) is better explained by a five-factor model rather than the four-factor model in the standardization sample. The current study examined the WISC-IV's factor structure in a sample of children with ADHD.

METHOD

Participants included 314 children and adolescents who were diagnosed with ADHD. Confirmatory factor analysis was conducted on the 10 core subtests of the WISC-IV, and three models were examined including two based on Cattell-Horn-Carroll (CHC) theory.

RESULTS

A five-factor model consisting of Gc, Gf, Gv, Gsm, and Gs factors provided the best fit for the data. The Perceptual Reasoning factor identified in the original four-factor model split into the two CHC factors, Gf and Gv, and cross-loaded the Symbol Search subtest onto the Gv factor.

CONCLUSION

A five-factor model based on CHC theory provided superior fit for the WISC-IV in children with ADHD, as has been found with the standardization sample.

Factor structure of symptom dimensions in attention-deficit/hyperactivity disorder (ADHD)

There is disagreement on whether attention-deficit/hyperactivity disorder (ADHD) symptoms are best characterized along two dimensions consisting of inattention and hyperactivity/impulsivity or three dimensions where hyperactivity and impulsivity are separate. To address this, the current study investigated the underlying symptom dimensions of ADHD by examining two- and three-factor models of ADHD symptom ratings in 400 children and adolescents diagnosed with ADHD. ADHD symptom ratings for each of the 18 DSM-IV Criteria A symptoms were obtained from mothers using a standardized symptom rating scale. Confirmatory factor analysis (CFA) was used to examine whether the 18 symptoms were best explained by two or three latent constructs. Results of the CFA demonstrated that a three-factor model was superior to a two-factor model. Findings support three distinct symptom dimensions that are consistent with previous research demonstrating unique clinical presentations of inattention, hyperactivity, and impulsivity. Differentiating between these three domains may aid in predicting behavioral outcomes in children with ADHD.

Psychometric properties of the Body Checking Questionnaire in college women

While the Body Checking Questionnaire (BCQ; Reas et al., 2002) is the most commonly-used measure of body checking behaviors, findings on the factor structure in nonclinical samples are mixed. This study investigated the factor structure and psychometric properties of the BCQ among nonclinical college women. In Study 1 (n=326), an exploratory factor analysis indicated factors corresponding to (a) behavioral and (b) visual checking. In Study 2 (n=1013), a confirmatory factor analysis revealed adequate fit for these factors, better than the original or total score solutions. Scales based upon the two factors demonstrated good internal consistency, convergent validity with measures of eating pathology, and sensitivity for detecting at-risk eating pathology. Results suggest an alternate two-factor solution that differs from the original three-factor solution.

Differences in developmental changes in academic and social premorbid adjustment between males and females with schizophrenia

The assessment of premorbid adjustment in schizophrenia has received considerable attention because of models suggesting that schizophrenia is a neurodevelopmental disorder characterized by abnormalities in functioning prior to onset of the disorder. Some studies suggest that premorbid adjustment is best viewed as a multidimensional construct where different areas of functioning might be differentially impacted by the illness and sex. The current study examined these matters using of Premorbid Adjustment Scale (PAS) in a sample of 421 individuals with schizophrenia. Confirmatory factory analyses conducted for three developmental periods (childhood, early adolescence, late adolescence) and for males and females separately, indicated that the PAS consists of academic and social factors that are invariant across developmental period and sex. However, differences in severity of academic and social premorbid impairment were present between males and females across developmental periods. Findings suggest important differences between males and females in the course of premorbid deterioration prior to onset of schizophrenia.

Measuring the ability to perceive the emotional connotations of written language

Emotionally laden writing is essential to our personal and professional lives. The purpose of this article was to design and evaluate a new test of the ability to decode the emotional connotations of written material. A series of 3 studies (totaling 457 participants) were used to demonstrate that the Metaphors Test measures a single construct, has strong internal consistency, has strong convergent validity with tests related to emotional and social intelligence, and has strong discriminant validity with vocabulary and personality. Confirmatory factor analyses demonstrated that the Metaphors Test is more closely associated with emotion perception than emotional understanding. Unlike most other tests that tap this skill, the stimuli for the Metaphors Test do not include any explicit emotion words; it is therefore a unique and valuable measure of emotion perception.

Factor structure of the Comprehensive Trail Making Test in children and adolescents with brain dysfunction

The Comprehensive Trail Making Test (CTMT) is a relatively new version of the Trail Making Test that has a number of appealing features, including a large normative sample that allows raw scores to be converted to standard T scores adjusted for age. Preliminary validity information suggests that CTMT scores are sensitive to brain injury and demonstrate expected correlations with other neuropsychological tests, although the evidence also suggests that the factor structure of the CTMT may differ in children with brain dysfunction in comparison to the standardization sample. The present study addresses this matter by conducting a confirmatory factor analysis (CFA) of the CTMT in 382 children and adolescents. Of the 382, 191 were diagnosed with various forms of brain dysfunction, including 140 who had sustained a traumatic brain injury. The other 191 participants were healthy controls who were individually matched on age and gender to the clinical group with brain dysfunction. Results of the CFA indicated that for the clinical group, a 2-factor model composed of Simple Sequencing and Complex Sequencing/Shifting factors provided the best fit for the data. Although these 2 factors were also identified in the controls, differences in pattern of loadings were present. Results suggest that the presence of brain dysfunction may alter the factor structure of the CTMT in children and adolescents.

Beery-Buktenica Developmental Test of Visual-Motor Integration performance in children with traumatic brain injury and attention-deficit/hyperactivity disorder

Evaluation of visuoconstructional abilities is a common part of clinical neuropsychological assessment, and the Beery-Buktenica Developmental Test of Visual-Motor Integration (VMI; K. E. Beery & N. A. Beery, 2004) is often used for this purpose. However, few studies have examined its psychometric properties when used to assess children and adolescents with traumatic brain injury (TBI) or attention-deficit/hyperactivity disorder (ADHD), even though these are among the most common acquired and neurodevelopmental forms of brain dysfunction in children. This study examined the validity of VMI scores in 123 children with TBI and 65 with ADHD. The TBI and ADHD groups performed significantly worse than the standardization sample, obtaining VMI mean scores of 87.2 (SD = 13.7) and 93.5 (SD = 11.27). Previous research has noted decrements in visuoconstructional abilities in TBI but relative sparing in ADHD. To examine the criterion validity of VMI scores, the authors therefore compared these 2 groups. As anticipated, the TBI group performed significantly worse than the ADHD group, but receiver operator characteristic analysis indicated that VMI scores were poor at discriminating between groups. Nonetheless, convergent validity evidence supported interpretation of VMI scores as measuring perceptual organization in both groups. In particular, principal components analysis indicated that VMI total scores loaded with perceptual organization tests from the Wechsler Intelligence Scale for Children, 3rd ed. (WISC-III; D. Wechsler, 1997), and its highest correlation was with the WISC-III Perceptual Organization Index. Also, the VMI correlated significantly with the Grooved Pegboard test for the group with TBI. These findings suggest that VMI scores are sensitive to visuoconstructional and motor deficits in children with developmental and acquired brain dysfunction.

False recognition of facial expressions of emotion: causes and implications

This article examines the importance of semantic processes in the recognition of emotional expressions, through a series of three studies on false recognition. The first study found a high frequency of false recognition of prototypical expressions of emotion when participants viewed slides and video clips of nonprototypical fearful and happy expressions. The second study tested whether semantic processes caused false recognition. The authors found that participants made significantly higher error rates when asked to detect expressions that corresponded to semantic labels than when asked to detect visual stimuli. Finally, given that previous research reported that false memories are less prevalent in younger children, the third study tested whether false recognition of prototypical expressions increased with age. The authors found that 67% of eight- to nine-year-old children reported nonpresent prototypical expressions of fear in a fearful context, but only 40% of 6- to 7-year-old children did so. Taken together, these three studies demonstrate the importance of semantic processes in the detection and categorization of prototypical emotional expressions.

Bias in consensus scoring, with examples from ability emotional intelligence tests

Consensus scoring occurs when the scoring key for a test is based upon the responses of the norm group. Consensus scoring is an attractive alternative to traditional methods of creating a scoring key for ability tests, especially useful when experts disagree about the correct answers to test items, as they do in the area of emotions and emotion perception. Of the many variations of consensus scoring, mode consensus scoring (the most frequent response in a norm group is given a score of 1, and all other responses a score of 0) and proportion consensus scoring (each respondent's score on an item is equal to the proportion of the norm group who match the respondent's answer) are the most widely used and the most psychometrically promising. This paper demonstrates that mode consensus scoring is biased against smaller sub-groups within the norm group: when sub-groups differ in their modal responses, the size of the sub-groups will influence the average group score. No known scoring option eliminates this bias. In contrast, proportion consensus scoring is not necessarily biased against smaller groups, although bias does occur in some extreme situations. Proportion consensus scoring is therefore the preferred consensus scoring option at this time.

Toward More Robust Inferential Procedures for Coefficient Alpha Under Sampling of Both Subjects and Conditions

Barchard and Hakstian (1997b) demonstrated that inferential procedures used with coefficient alpha are not robust under sampling of both subjects and conditions (Type 12 sampling) with measurement data departing from essentially-parallel form. In the first of 2 studies reported here, a sample-based, non-analytical degrees-of-freedom correction factor was empirically developed that correlated almost perfectly with the independently-established correct degrees of freedom for the data. In the second study, application of this correction factor was assessed through a comprehensive simulation study involving Type 12 data sets representing a wide range of design characteristics and manifesting tau-equivalent measurement. Use of the correction factor yielded actual Type I error rates closer to nominal values than were obtained using uncorrected inferential procedures. Implications for practice and future research are discussed.

The Robustness of Confidence Intervals for Coefficient Alpha Under Violation of the Assumption of Essential Parallelism

Two methods exist for setting confidence intervals for coefficient alpha: Feldt's (1965) exact F-distribution method and Hakstian and Whalen's (1976) asymptotic normal-theory procedure. These methods both assume that items are essentially parallel. Two studies, employing Type 12 sampling, are reported in which the effects of violating this assumption are examined. The results indicate that as long as the data manifest the properties of essential parallelism, the two methods maintain precise Type 1 error control. With a variety of forms of measurement data containing correlated errors or unequal variances, however, the methods resulted in poor Type I error control. No performance differences between the two methods were found. Implications are noted for practice and for future research on inference with coefficient alpha.