Reexamining the factor structure of the 20-item Toronto alexithymia scale: commentary on Gignac, Palmer, and Stough

Gignac, Palmer, and Stough (2007/this issue) test a number of different latent factor models for the TAS-20 using a community sample of 355 participants and conclude that this scale is best represented by a "nested factors model," with five substantive factors and a method factor. Gignac et al. also report that the correlated three-factor model and a comparable higher order model supported by most studies produced poor levels of incremental close fit. In this article, we challenge Gignac et al.'s unheralded and largely unsupported use of nested model fitting and the uncritical acceptance of exceptionally high cutoff levels to assess goodness of fit (GOF). Using more traditional and empirically supported model testing procedures and a more flexible approach to the interpretation of multiple tests of GOF, we interpret Gignac et al.'s results as actually supportive of the traditional three-factor model and one that has been recovered in 17 of the 24 factor analytic studies of the TAS-20.     

The Prospective and Retrospective Memory Questionnaire (PRMQ): factorial structure, relations to global subjective memory ratings, and Swedish norms

The factorial structure of the Prospective and Retrospective Memory Questionnaire (PRMQ; Smith et al., 2000) was examined in a Swedish population based sample (N= 540, age range; 35-90 years). Concurrent validity was assessed by relating PRMQ to global ratings of memory. Confirmatory factor analyses of the PRMQ items indicated a superior fit of a three-factor model, with prospective and retrospective memory as orthogonal factors and episodic memory as a common factor. Furthermore, the PRMQ scales correlated with the global ratings of memory, suggesting that each rating contributed with unique variance in predicting PRMQ scores. Given differences in levels of complaints as compared with prior research (Crawford et al., 2003) norms for the Swedish version are provided. In conclusion, the present findings extend earlier work by providing additional support for the construct and concurrent validity of the PRMQ scales.     

Factor structure of the Behavior Rating Inventory of Executive Functions (BRIEF-P) at age three years

The preschool period is an important developmental period for the emergence of cognitive self-regulatory skills or executive functions (EF). To date, evidence regarding the structure of EF in preschool children has supported both unitary and multicomponent models. The aim of the present study was to test the factor structure of early EF as measured by the Behavior Rating Inventory of Executive Function-Preschool version (BRIEF-P). BRIEF-P consists of five subscales and three broader indexes, hypothesized to tap into different subcomponents of EF. Parent ratings of EF from a nonreferred sample of children recruited from the Norwegian Mother and Child Cohort Study (N = 1134; age range 37–47 months) were subjected to confirmatory factor analyses (CFA). Three theoretically derived models were assessed; the second-order three-factor model originally proposed by the BRIEF-P authors, a “true” first-order one-factor model and a second-order one-factor model. CFA fit statistics supported the original three-factor solution. However, the difference in fit was marginal between this model and the second-order one-factor model. A follow-up exploratory factor analysis (EFA) supported the existence of several factors underlying EF in early preschool years, with a considerable overlap with the five BRIEF-P subscales. Our results suggest that some differentiation in EF has taken place at age 3 years, which is reflected in behavior ratings. The internal consistency of the BRIEF-P five clinical subscales is supported. Subscale interrelations may, however, differ at this age from those observed in the preschool group as a whole.     

An exploratory investigation of the reliability and validity of the Independent-Interdependent Problem-Solving Style Scale

The Independent-Interdependent Problem-Solving Scale is based on Cross et al.'s conceptualisation of relational-interdependent self-construal. The IIPSS provides a relatively context-free measure of people's tendencies to solve problems independently or with the help of others. Because previous investigations have not provided extensive evidence for the reliability and validity of the IIPSS, the current research aimed to test the psychometric properties of this novel measure. Investigations of four student samples (combined N = 1157) and one sample comprised of academic researchers (N = 198) generally supported the reliability and validity of the IIPSS. Exploratory factor analysis of IIPSS items yielded a single factor structure. However, confirmatory factor analyses did not demonstrate good model fit for the one factor solution and instead yielded good model fit for two underlying factors. The IIPSS showed adequate test–retest reliability and predicted positive associations with social personality traits. It also showed no significant associations with measures of demand characteristics and social desirability. Future research needs to be undertaken to further assess the factor structure and address shortcomings of the present research such as utilising objective data in addition to self-reports to assess the scale's validity.     

A modification of the Fenigstein, Scheier, and Buss Self-Consciousness Scales

Confirmatory factor analysis was employed to examine the structure of the Fenigstein, Scheier, and Buss (1975) Self-consciousness Scales. Several weaknesses were identified and modifications were made to improve the representation of the underlying structure of the data. Five items were found to be relatively unreliable and were dropped. Two separate but correlated private self-consciousness dimensions, self-reflectiveness and internal state awareness, were identified. The resulting four-factor self-consciousness model fit the data more adequately than the three-factor model of Fenigstein et al.     

Exploration of the Factor Structure of ADHD in Adolescence through Self,Parent, and Teacher Reports of Symptomatology

Factor analytic studies of attention-deficit/hyperactivity disorder (ADHD) in children and adults have shown that second-order and bifactor models better represent ADHD symptoms than two- or three-factor models, yet there is far less evidence for a bestfitting model of ADHD in adolescence. Thus, the current study examined the factor structure of ADHD in adolescence and further evaluated the external validity of the best fitting model. Participants were 588 adolescents (22 % female; 366 with a childhood ADHD diagnosis; mean age 15.9 years) from the 8-year assessment of the Multimodal Treatment Study of Children with ADHD (MTA). ADHD symptoms were assessed via adolescent self-report, parent report, and teacher report on the SNAP-IV scale. Potential factor structures for the 18 symptoms of ADHD were tested for each informant, which included traditional one-factor, two-factor, and three-factor models of ADHD, as well as second-order factor (specific factors loading onto general factor) and bifactor (items loading onto both specific and general factors) models. Unique associations between external criteria and the identified factors of each informant’s best fitting model were examined. Although several of the proposed models exhibited good fit, the second-order two-factor model best accounted for ADHD in adolescence according to self-report and parent report, and the second-order three-factor model was optimal according to teacher report. Several key measurement issues emerged for the hierarchical bifactor models, such as numerous Heywood cases and out-of-bound parameter estimates, which rendered them unfit as optimal representations of ADHD in adolescence. These findings and the implications of the best fitting model of ADHD in adolescence suggest that a possible reorganization of this disorder may eventually aid clinicians in the accurate diagnosis of ADHD in adolescents.     

Confirmatory analysis of simultaneous,sequential, and achievement factors on the K-ABC at 11 age levels ranging from 212 to 1212 years

The K-ABC is a recently published measure of children's intelligence based on a theoretical model of information processing. Exploratory factor analyses conducted during the scale's development supported the underlying theoretical model of sequential, simultaneous, and achievement factors. Confirmatory analysis using a Jöreskog method was carried out in the present investigation and completely supported the two-factor processing model. When the achievement measures were included, the two mental processing factors again clearly emerged, along with a third factor defined by the achievement scale subtests, but they did not fit an independent three-factor model.     

Alexithymia and its measurement: confirmatory factor analyses of the 20‐item Toronto Alexithymia Scale and the Bermond–Vorst Alexithymia Questionnaire

Alexithymia refers to the difficulties an individual has in experiencing and expressing feelings. The 20‐item Toronto Alexithymia Scale (TAS‐20) and the Bermond–Vorst Alexithymia Questionnaire (BVAQ) with two parallel versions of 20 items have been constructed to measure it. The present study compared the psychometric properties of these two self‐report questionnaires in English (N=290) and French (N=322) language versions. Confirmatory factor analysis was used to examine the fit between the hypothesized factors and the data. Results revealed a better fit to the data for the second version of the BVAQ (BVAQ‐20B) for both language versions as compared to the TAS‐20, the whole BVAQ, or the first version of the BVAQ (BVAQ‐20A). Additionally, the factor comparison of both language versions indicated that only the factorial structure of the BVAQ‐20B was replicable across languages. Concurrent validity of the questionnaires is discussed. Copyright © 1999 John Wiley & Sons, Ltd.     

Investigation of the factorial structure and psychometric properties of the Persian version of the Toronto Alexithymia Scale-20 in patients with psychiatric disorders

The aims of this study were to assess the factor structure, validity, and reliability of the Persian translation of the Toronto Alexithymia Scale-20 (TAS-20) and to examine different models of the TAS-20 in Iranian patients with various psychiatric disorders. Participants were 839 patients with psychiatric disorders, including obsessive-compulsive disorder (OCD) (n = 80), schizophrenia (n = 82), bipolar disorder (BD) (n = 100), alcohol dependence (n = 81), major depressive disorder (MDD) (n = 95), psychosomatic disorders (n = 92), anxiety disorders (n = 85), post-traumatic stress disorder (PTSD) (n = 90), attention deficit hyperactivity disorder (ADHD) (n = 55), and suicide attempts (n = 79). Results indicated that the three-factor TAS-16 fit the data well, after removing four items from the externally-oriented thinking (EOT) subscale. In addition, the total score and subscales had strong internal consistency and concurrent validity. An alternative three-factor model and a four-factor model, which both allow the reverse-coded EOT items to load on a separate factor, also had an acceptable fit. The results suggest that after deleting four items from the EOT subscale, the 16-item TAS is a reliable scale among Iranian psychiatric patients. Moreover, the alternative three-factor and four-factor structures may be appropriate to apply among Iranian patients.

     

Construct Validity of the Acquired Capability for Suicide Scale: Factor Structure,Convergent and Discriminant Validity

The Acquired Capability for Suicide Scale (ACSS) assesses one of three main constructs in the Interpersonal Theory of Suicide but evidence of its validity is limited. In two studies (Ns = 287 and 738) validity of the full 20-item ACSS and its shorter versions (ACSS-5, ACSS-8, ACSS-FAD) were examined in terms of factor structure and relation to indices of self-reported suicidal behaviour and self-harm. Confirmatory factor analysis (CFA) failed to show good fit for one-, two-, or three-factor models of the ACSS in its various versions. Exploratory factor analysis of the 20-item scale in the first study pointed instead to a five-factor structure and this was supported using CFA in the second study. In both studies all scale versions showed moderate negative correlations to fear of death and dying, indicating scale validity for the purpose of assessing fearlessness about death. In the second study, a model in which the five factors were indicators of a latent variable of Capability was found to predict a latent variable of Suicidality as indicated by suicidal behaviours, but the prediction was substantially enhanced by the addition of Item 20 to the model. This single item was also found in the first study to better predict suicidal and self-harming behaviour than the full ACSS or any of its short versions.     

Effects of divided attention in the word-fragment completion task with unique and multiple solutions

A study with 114 young adults investigated the correlations of intelligence factors and working-memory capacity with reaction time (RT) tasks. Within two sets of four-choice RT tasks, stimulus–response compatibility was varied over three levels: compatible, incompatible, and arbitrary mappings. Two satisfactory measurement models for the RTs could be established: A general factor model without constraints on the loadings and a nested model with two correlated factors, distinguishing compatible from arbitrary mappings, with constraints on the loadings. Structural models additionally including factors for working memory and intelligence showed that the nested model with correlated factors is superior in fit. Working-memory capacity and fluid intelligence were correlated strongly with the nested factor for the RT tasks with arbitrary mappings, and less with the general RT factor. The results support the hypothesis that working memory is needed to maintain arbitrary bindings between stimulus representations and response representations, and this could explain the correlation of working-memory capacity with speed in choice RT tasks.     

The Brief Problem Monitor (BPM-Y/BPM-P) Among Chinese Youth: Psychometric Properties and Measurement Invariance

The Brief Problem Monitor (BPM) is a 19-item instrument newly developed for screening children and adolescents for emotional and behavioral problems. The aim of this study was to investigate the psychometric properties of the youth-report and parent-report versions of the BPM (BPM-Y and BPM-P) among Chinese youth. Data was collected from two samples: a middle school student sample (n?=?1,246, girls?=?516 [41.4%]) and a community youth sample from families suffering adversity (n?=?383, girls?=?143 [37.3%]). We conducted a series of confirmatory factor analyses to test and compare the latent structures of the BPM-Y and BPM-P. Results suggested that the three-factor model of the BPM-Y and BPM-P fit the data best. And based on the omnibus fit indices and difference between them, the items of the BPM-Y and BPM-P support measurement invariance across informants (both youth- and parent-report) for the three-factor model. Internal consistency coefficients were satisfactory (Cronbach's α?=?.73?~?.90). In terms of convergent validity, BPM factors and BPM scores showed expected correlations with strength and difficulties behavior, callous–unemotional traits, psychopathic traits, and social competence. In sum, our findings supported that the three-factor structuure of BPM-Y and BPM-P could be a promising instrument to screen emotional and behavioral problems in Chinese youth.

     

A Bifactor Model of Anxiety Sensitivity: Analysis of the Anxiety Sensitivity Index-3

The purpose of the current study was to investigate the fit of a bifactor model of the Anxiety Sensitivity Index-3 (ASI-3; Taylor et al. Psychological Assessment, 19, 176–188, 2007) as well as to examine measurement invariance of the ASI-3 across gender. Participants were undergraduate students from the University of Cincinnati (n?=?954; 63 % female). Results indicated that the bifactor model was the best fit to the data, and that this model of AS is fully invariant in terms of gender. The current findings suggest that anxiety sensitivity consists of a general factor and three independent group factors (rather than a higher-order factor with three correlated lower-order factors). The ASI-3 subscales from the bifactor model however did not provide incremental predictive utility above and beyond the general AS factor with respect to an external anxiety criterion. Related research and clinical implications are discussed.     

The structure of the Hospital Anxiety and Depression Scale: Theoretical and methodological considerations

The Hospital Anxiety and Depression Scale (HADS; Zigmond - Snaith, 1983) is widely used; however, its factor structure is unclear, with studies reporting differing unidimensional, two-factor and three-factor models. We aimed to address some key theoretical and methodological issues contributing to inconsistencies in HADS structures across samples. We reviewed existing HADS models and compared their fit using confirmatory factor analysis (CFA). We also investigated methodological effects by comparing factor structures derived from Rasch and Principal Components Analysis (PCA) methods, as well as effects of a negative wording factor. An Australian community-dwelling sample consisting of 189 females and 158 males aged 17–86 (M = 35.73, SD = 17.41) completed the 14-item HADS. The Rasch Analysis, PCA and CFA all supported the original two-factor structure. Although some three-factor models had good fit, they had unacceptable reliability. In the CFA, a hierarchical bifactor model with a general distress factor and uncorrelated depression and anxiety subscales produced the best fit, but the general factor was not unidimensional. The addition of a negative wording factor improved model fit. These findings highlight the effects of differing methodologies in producing inconsistent HADS factor structures across studies. Further replication of model fit across samples and refinement of the HADS items is warranted.     

Confirmatory factor analysis of the eating disorder inventory

The Japanese version of the Eating Disorder Inventory was completed by 204 female university students. Confirmatory factor analyses supported the 9 factors reported by Shimura, et al. (2003). Further, maximum likelihood factor analysis yielded nine factors, accounting for 48.8% of the total variance. There was also adequate fit for each of the 9 subscales. This study provided additional evidence on validity for this Japanese version.     

Competitive anxiety in young athletes: differentiating somatic anxiety, worry, and concentration disruption

The age-appropriate Sport Anxiety Scale-2 (SAS-2; Smith, Smoll, Cumming, & Grossbard, 2006) was used to assess levels of cognitive and somatic anxiety among male and female youth sport participants. Confirmatory factor analyses with a sample of 9-14 year old athletes (N=1038) supported the viability of a three-factor model of anxiety involving somatic anxiety, worry, and concentration disruption previously demonstrated in high school and college samples. Tests for factorial invariance revealed that the three-factor model was an equally good fit for 9-11 year olds and 12-14 year olds, and for both males and females. Gender and age were modestly related to anxiety scores. Worry about performing poorly was highest in girls and in older athletes, whereas boys reported higher levels of concentration disruption in competitive sport situations. Implications for emotional perception and for the study of competitive anxiety in young athletes are discussed.     

Model Fit and Item Factor Analysis: Overfactoring,Underfactoring, and a Program to Guide Interpretation

In exploratory item factor analysis (IFA), researchers may use model fit statistics and commonly invoked fit thresholds to help determine the dimensionality of an assessment. However, these indices and thresholds may mislead as they were developed in a confirmatory framework for models with continuous, not categorical, indicators. The present study used Monte Carlo simulation methods to investigate the ability of popular model fit statistics (chi-square, root mean square error of approximation, the comparative fit index, and the Tucker–Lewis index) and their standard cutoff values to detect the optimal number of latent dimensions underlying sets of dichotomous items. Models were fit to data generated from three-factor population structures that varied in factor loading magnitude, factor intercorrelation magnitude, number of indicators, and whether cross loadings or minor factors were included. The effectiveness of the thresholds varied across fit statistics, and was conditional on many features of the underlying model. Together, results suggest that conventional fit thresholds offer questionable utility in the context of IFA.     

Latent mean comparisons on the BRIEF in preterm children: Parent and teacher differences

Executive function is a heterogeneous construct applied to cognitive capacities that together enable individuals to effectively engage in activities toward a purposive goal. Children born at extremely low birth weight (ELBW) are at risk of executive dysfunction on performance-based measures. In natural contexts, executive function may be described using such parental and teacher questionnaires as the Behavior Rating Inventory of Executive Function (BRIEF). This study examined the factor structure of the BRIEF-parent form in 124 ELBW children and of the BRIEF-teacher form in 90 ELBW children. Although our data showed that the fit of a two-factor structure was adequate for the parent report, a three-factor model provided advantages over the two-factor model across all fit indices and best characterized the data. For teacher report, these data supported a three-factor but not a two-factor model. Using the three-factor model for both groups of informants, we compared parent and teacher reports (n = 90 pairs) between the three identified latent variables. Parents reported significantly more difficulty with Emotional Regulation (p < .05), and teachers reported significantly more difficulty with Behavioral Regulation (p < .05). No significant differences were found between parent and teacher reports for Metacognition.     

Psychometric properties of a revised Spanish 20-item Toronto Alexithymia Scale adaptation in multiple sclerosis patients

There have been a small number of investigations of alexithymia in multiple sclerosis (MS) using the 20-item Toronto Alexithymia Scale (TAS-20). However, the TAS-20 factor structure has not yet been evaluated in a MS patient sample, and earlier Spanish translations of this instrument require some improvement. We aimed to evaluate the factorial validity and reliability of an improved Spanish translation of the TAS-20 (the TAS-20-S). The TAS-20-S was completed by 221 MS patients. Confirmatory factor analysis was used to compare the fit of six different factor models. Internal consistency and retest reliability coefficients were also computed. The correlated three-factor model and the higher-order factor model made up of Difficulty Identifying Feelings, Difficulty Describing Feelings, and Externally Oriented Thinking achieved the best fit. Alpha coefficients ranged between .87 and .67; mean inter-item correlations ranged between .48 and .20; and retest correlations after 6 months ranged between .61 and .52. A high degree of alexithymia was present in 18.1% of the sample. Reliability and the traditional three-factor structure were demonstrated for the TAS-20-S, which can now be recommended for assessing an aspect of emotional processing in MS patients.     

Confirmatory Factor Analysis of the Structure of Adult ADHD Symptoms

Previous research has examined the factor structure of ADHD symptoms in adults using confirmatory factor analysis and there is mixed support for two-factor, three-factor, or more complex bifactor models. The current study re-examines these factor structures on the basis of not only overall model fit, but also criteria such as chi-square difference tests, invariance, reliability, and validity. Three datasets were pooled to form a total sample of 430 adults (50% female). Participants completed questionnaires assessing their ADHD symptoms, education, psychopathology, and parenting (the latter as validity variables). Although the bifactor models demonstrated the best fit, the reliability and construct replicability of specific factors in these models was poor. The three-factor model had good fit and demonstrated good reliability, validity, construct replicability, and some invariance across gender and dataset. Our results replicate previous studies that find superior fit for bifactor models. However, the superior statistical fit may reflect only the increased complexity of these models. The more parsimonious three-factor model may demonstrate stronger validity and clinical utility.