The Child Concentration Inventory (CCI): Initial ...

CHILD CONCENTRATION INVENTORY This article may not exactly replicate the final version published in the APA journal. It is not the copy of record. Here is the link to the final version published by APA: http://psycnet.apa.org/psycinfo/2015-04303-001/ Becker, S. P., Luebbe, A. M., & Joyce, A. M. (2015). The Child Concentration Inventory (CCI): Initial validation of a child self-report measure of sluggish cognitive tempo. Psychological Assessment. Advance online publication. doi: 10.1037/pas0000083

The Child Concentration Inventory (CCI): Initial Validation of a Child Self-Report Measure of Sluggish Cognitive Tempo

Stephen P. Becker1 Aaron M. Luebbe2 Ann Marie Joyce3,4

1

Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center,

Cincinnati, Ohio, USA 2

Department of Psychology, Miami University, Oxford, Ohio, USA

3

Community Mental Health, IWK Health Centre, Halifax, Nova Scotia, Canada

4

Department of Psychology and Neuroscience, Dalhousie University, Halifax, Nova Scotia, Canada

Author note: Stephen P. Becker and Aaron M. Luebbe contributed equally to this article and share first authorship. The measure described in this manuscript is an adaptation of the sluggish cognitive tempo scale developed by Ann Marie Joyce (nee Penny) and colleagues (“Developing a measure of sluggish cognitive tempo in children: Content validity, factor structure, and reliability”, 2009, Psychological Assessment, 21, pp. 380-389). Correspondence concerning this article should be addressed to Stephen P. Becker, Division of Behavioral Medicine and Clinical Psychology, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, MLC 10006, Cincinnati, Ohio 45229-3039; (513) 803-2066 (phone); (513) 803-0084 (fax); [email protected] (e-mail).

1

CHILD CONCENTRATION INVENTORY

2

Abstract Sluggish cognitive tempo (SCT) is characterized by excessive daydreaming, mental confusion, slowness, and low motivation. Several teacher- and parent-report measures of SCT have recently been developed but a child self-report measure of SCT does not yet exist despite clear links between SCT and internalizing psychopathology (for which self-report is often desired). This study examined the initial reliability and validity of the Child Concentration Inventory (CCI), a child self-report measure of SCT symptoms, in a school-based sample of 124 children (ages 8-13; 55% female). Children completed the CCI and measures of academic/social functioning, emotion regulation, and self-esteem. Teachers completed measures of psychopathology symptoms (including SCT) and academic/social functioning. Although exploratory structural equation modeling (ESEM) supported a three-factor model of the CCI (consisting of Slow, Sleepy, and Daydreamer scales closely resembling the factor structure of the parent-report version of this measure), bifactor modeling and omega reliability indices indicated that the CCI is best conceptualized as unidimensional. CCI scores were significantly correlated with teacher-rated SCT and were statistically distinct from teacher-rated ADHD and child-rated anxiety/depression. After controlling for sex, grade, and other psychopathology symptoms, the CCI total score was significantly associated with poorer childreported academic/social functioning and self-worth in addition to increased loneliness and emotion dysregulation. Child ratings on the CCI were moderately-to-strongly correlated with poorer teacherrated academic/social functioning but these associations were reduced to nonsignificance after controlling for demographics and other psychopathology symptoms. Findings provide preliminary support for the CCI, and future directions include replication with adolescents and clinical samples in order to further examine the CCI’s factor structure, reliability, validity, and clinical utility. Key Words: ADHD; attention-deficit/hyperactivity disorder; concentration deficit disorder; factor structure; functional impairment


3

The Child Concentration Inventory (CCI): Initial Validation of a Child Self-Report Measure of Sluggish Cognitive Tempo Since the publication of several seminal articles over a decade ago (Carlson & Mann, 2002; McBurnett, Pfiffner, & Frick, 2001; Milich, Balentine, & Lynam, 2001), there has been a marked increase in studies examining the construct of Sluggish Cognitive Tempo (SCT; see Becker, Marshall, & McBurnett, 2014, for a review). SCT comprises symptoms of excessive daydreaming, mental confusion or “fogginess,” slowness, and low motivation/initiative. These symptoms were initially identified as potentially useful for identifying a subset of children with Attention-Deficit/Hyperactivity Disorder (ADHD) who showed clinical levels of inattentive symptoms but few if any symptoms of hyperactivity-impulsivity (Barkley, DuPaul, & McMurray, 1990; Lahey, Schaughency, Frame, & Strauss, 1985). However, SCT symptoms have never been included as ADHD symptoms in the Diagnostic and Statistical Manual of Mental Disorders (DSM), in part because SCT symptoms showed poor negative predictive power for predicting a diagnosis of ADHD in the DSM-IV field trials (Frick et al., 1994). Although there continues to be interest in whether the co-occurrence of SCT is associated with unique impairments in children with ADHD (Capdevila-Brophy et al., 2012; Marshall et al., 2014; Willcutt et al., 2014), there is a convergence of research demonstrating SCT to be a distinct construct from ADHD (Becker, 2013). In support of this possibility, over a dozen studies conducted with children (Barkley, 2013; Becker, Luebbe, Fite, Stoppelbbein, & Greening, 2014; Lee et al., 2014; Willcutt et al., 2014) and adults (Barkley, 2012; Becker, Langberg, Luebbe, Dvorsky, & Flannery, 2014) show SCT symptoms to be statistically distinct from both ADHD inattention and hyperactivity-impulsivity. Bifactor modeling has also supported the distinctiveness of SCT from ADHD (Garner et al., 2014).


4

As research has consistently shown SCT to be distinct from ADHD, there has been significant interest in uncovering the developmental and clinical correlates associated with SCT in order to further understand the construct and determine its external validity. In order for this work to proceed, there was until recently a dire need for better measures of SCT. A wellvalidated measure of SCT symptoms did not exist until 2009, when Penny and colleagues (Penny, Waschbusch, Klein, Corkum, & Eskes, 2009) developed a parent- and teacher-report SCT scale. Prior to this time, most studies of SCT relied on using a handful of SCT items that had been somewhat serendipitously included on frequently-used measures such as the Child Behavior Checklist (CBCL) or Teacher’s Report Form (TRF; Achenbach & Rescorla, 2001). As such, the validation of the Penny et al. (2009) measure of SCT offered a much-needed contribution to the field, and several other parent- and teacher-report measures of SCT have since been developed (Barkley, 2013; Lee et al., 2014; McBurnett et al., 2014). An adult selfreport measure of SCT (based largely on the Penny et al., 2009 measure) has also been published (Barkley, 2012). To date, however, a validated child self-report measure of SCT does not exist, leaving a critical gap in the options available to researchers seeking to gain a fuller understanding of the SCT construct or collect ratings of SCT symptoms from multiple informants. Rationale for a Child Self-Report Measure of SCT There are several reasons why a self-report measure of SCT is needed. First, it is wellestablished that discrepancies exist among informants of children’s psychopathology (Achenbach, McConaughy, & Howell, 1987; De Los Reyes & Kazdin, 2006). Correlations among parent, teacher, and child ratings of children’s psychopathology are often modest at best, making multi-informant assessment important for both research and clinical purposes. Since a


5

validated child self-report measure of SCT does not exist, the degree to which children’s ratings of SCT symptoms correlate with teacher or parent ratings is entirely unknown, and it is likewise unknown whether child-rated SCT symptoms are uniquely associated with children’s psychosocial adjustment. Second, one of the most consistent findings in the SCT literature is that of a significant positive association between SCT symptoms and internalizing symptoms of anxiety and depression, an association that remains after controlling for ADHD symptom severity (Becker, Langberg et al., 2014; Becker, Luebbe et al., 2014; Becker & Langberg, 2013; Burns et al., 2013; Lee et al., 2014; Penny et al., 2009). Several studies indicate that SCT may be more strongly associated with depression than with anxiety (Barkley, 2013; Becker, Luebbe et al., 2014; Cortés et al,. 2014; Jacobson et al., 2012). Importantly, despite an association between SCT and internalizing symptoms, factor analytic studies indicate that SCT is distinct from both depression and anxiety (Becker, Luebbe et al., 2014; Lee et al., 2014; Willcutt et al., 2014). Since it is often important to collect children’s self-ratings of internalizing symptomatology, it is important to have a child self-report measure of SCT available in order to examine the degree to which childreported SCT symptoms are associated with child-reported anxiety and depression and also to test whether SCT remains statistically distinct from anxiety and depression when these constructs are all measured using child-report measures. Such evidence would provide further support for SCT as a construct that is correlated with, but distinct from, anxiety and depression. Third, and related to the two previous points, there is evidence suggesting SCT to more closely align with the internalizing versus externalizing spectrum of psychopathology. Specifically, Becker and colleagues (2013) examined personality dimensions of reward sensitivity and punishment sensitivity in relation to SCT and ADHD symptoms in a community


6

sample of school-aged children. As expected, sensitivity to reward (and impulsivity/fun-seeking specifically) was significantly associated with both general externalizing symptoms and ADHD symptoms specifically. In contrast, sensitivity to punishment (and fear/shyness specifically) was significantly associated with both general internalizing symptoms and SCT symptoms specifically. Although SCT has historically been studied within the context of ADHD, the findings of Becker et al. (2013) provide preliminary evidence that SCT may ultimately fall under the internalizing umbrella of psychopathology. If this is indeed the case, it is especially important to have a child self-report measure of SCT symptoms since children may be more valid reporters of their internalizing symptoms in comparison to their externalizing symptoms (Silverman & Ollendick, 2005). With these considerations in mind, the purpose of the present study was to develop and provide initial validation of a child self-report measure of SCT. What’s in a Name? Although the term “sluggish cognitive tempo” (or its variants such as “slow tempo”) has been in use for the last two decades (Becker, Marshall, & McBurnett, 2014), concerns have been raised that the term is both theoretically misleading and derogatory to patients and their families (Barkley, 2014). In terms of the first concern, an underlying cognitive deficit of SCT has not yet been identified, making the continued use of the term “sluggish cognitive tempo” akin to putting the terminological horse before the theoretical cart. Second, as noted by Barkley (2014), the term SCT may be off-putting and perceived as pejorative to patients, “smacking as it does of connotations of mental slowness, slow learning or wittedness, or frank mental retardation” (p. 123). In advocating for a focus on the attentional component of SCT (as distinct from ADHD) while also eliminating offensive connotations and avoiding the implication of an as-of-yet underterminded underlying dysfunction, Barkley (2014) recommended a change to the term


7

Concentration Deficit Disorder. We do not believe there is currently enough evidence to support use of the label “disorder” to describe the set of SCT symptoms (and SCT is not currently included in any diagnostic system), but we nonetheless share with Barkley the overarching concerns he has with the SCT label, and in response we have chosen to name our measure the Child Concentration Inventory (CCI), thus avoiding use of the SCT term in its name. The Present Study A validated child self-report measure of SCT symptoms does not currently exist, even though such a measure would advance the literature on the construct and also allow for interesting cross-reporter and predictive validity analyses. The aim of the present study was to develop and provide initial validation of the CCI, a child self-report measure of SCT. Specifically, we modified the Penny et al. (2009) parent- and teacher-report measure of SCT for use with children. We then administered this measure to a sample of school-aged children (grades 3-6) in order to examine the reliability, factor structure, construct validity, and criterion validity of this newly-created measure. Methods Participants The current study included 124 children (56 boys, 68 girls) attending an elementary school in the Midwestern United States. Children included in this study were students in third through sixth grades (ages 8-13; M = 10.51, SD = 1.30). According to official school records, and consistent with demographics of the surrounding community (95% White in the 2010 United States Census), the majority of participants in this study were White (n = 119; 96%) with remaining participants African American (n = 4; 3%) or Asian (n = 1; 1%). According to the 2010 Census, 28.4% of the city population was below the federal poverty level (median


8

household income = $30,299). Fifty-two percent (n = 65) of the students included in this study received free or reduced lunch. To further describe the sample, the county in which the school resides is classified by the 2013 Rural-Urban Continuum Codes as nonmetropolitan. Measures SCT Scale. Teachers completed the 14-item SCT Scale developed by Penny and colleagues (2009). In developing their measure, Penny and colleagues (2009) conducted a review of the literature in order to identify an initial SCT item pool and then used a group of experts in the field of SCT to evaluate the content validity of these items. This process led to the identification of the 14-item SCT Scale which was then validated in a sample of 335 elementary school-aged children. Each item on the SCT Scale is rated on a four-point scale (0 = not at all, 1 = just a little, 2 = pretty much, 3 = very much). Penny et al. (2009) provided support for the reliability (i.e., internal consistency, interrater reliability, and test-retest reliability) and external validity (i.e., convergent and discriminant validity) of scores from the SCT Scale. In a principal components analysis of the teacher-report version of the scale, a two-factor structure of SCT emerged, labeled as Sleepy/Daydreamer (e.g., “appears to be sluggish,” “seems drowsy,” “daydreams”, “seems to be in a world of his or her own”) and Slow (e.g., “is slow or delayed in completing tasks”, “lacks initiative to complete work,” “effort on tasks fades quickly”). However, three of the items cross-loaded on both factors and so a mean score of all 14 items was used in the present study for teacher-reported SCT (α = .93). Child Concentration Inventory (CCI). The Child Concentration Inventory (CCI) was developed for this study as a child self-report measure of SCT. The CCI was created by adapting the above-described SCT Scale (Penny et al., 2009). Specifically, each of the 14 items were changed to reflect the first person, and although we aimed to keep the items as similar as possible


9

to the original parent/teacher-report version so that versions could be comparable across respondents, the wording of several items was modified to make the items more easily understood by children. Table 1 of the supplemental materials lists the 14 items from the original adult-report SCT rating scale (Penny et al., 2009) alongside the revised items for the child-report CCI. Children rated each item on the same four-point scale of the Penny et al. SCT Scale (0 = not at all, 1 = just a little, 2 = pretty much, 3 = very much). Vanderbilt ADHD Diagnostic Teacher Rating Scale (VADTRS). In order to assess teacher-report of children’s ADHD, externalizing, and internalizing symptoms, teachers completed the 35-item Vanderbilt ADHD Diagnostic Teacher Rating Scale (VADTRS; Wolraich et al., 1998, 2013). The VADTRS includes 18 items that correspond to the DSM-IV symptoms of ADHD (9 inattentive items and 9 hyperactive-impulsive items) in addition to 10 oppositionaldefiant/conduct problem (ODD/CP) items and 7 anxiety/depression items. Each item is rated on a four-point scale (0 = never, 1 = occasionally, 2 = often, 3 = very often). Construct and convergent validity of the VADTRS scores have been established and the VADTRS subscale scores demonstrate acceptable internal consistency (Wolraich et al., 1998, 2013). As with previous research (Becker et al., 2012), separate anxiety (3 items) and depression (4 items) subscales were used in this study to increase specificity and to also allow us to examine whether child-reported CCI symptoms were more strongly associated with anxiety or depression. In the present study, internal consistencies were: ADHD Inattention α = .97, ADHD HyperactivityImpulsivity α = .90, ODD/CP α = .90, Anxiety α = .78, Depression α = .79. Revised Child Anxiety and Depression Scale – Short Version (RCADS-SV). The Revised Child Anxiety and Depression Scale – Short Version (RCADS-SV; Ebesutani, Reise et al., 2012) was used to assess for children’s self-reported depression and anxiety. The RCADS-


10

SV includes 10 depression items (e.g., “I feel sad or empty,” “nothing is much fun anymore”), and the 3 items representing generalized anxiety symptoms used in the present study (e.g., “I worry that something bad will happen to me”). Each item is rated on a four-point scale (0 = never, 1 = sometimes, 2 = often, 3 = always). Previous research demonstrates the reliability of scores from the RCADS-SV in school- and clinic-based samples (Ebesutani, Reise et al., 2012). In the present study, sum scale scores had adequate internal consistency (depression α = .80; generalized anxiety α = .72). Self-Perception Profile for Children (SPPC). Children and teachers both completed their respective versions of the Self-Perception Profile for Children (SPPC; Harter, 1985), a commonly-used measure of self- and other-perceived competence. Using a “some kids”/“other kids” format, ratings are made on a four-point scale, with higher scores indicating greater competence. The present study used the academic competence (child α = .82, teacher α = .96) and social acceptance (child α = .78, teacher α = .96) scales. Both of these domains have 6 and 3 items for the child and teacher versions, respectively. Children also completed the 6-item global self-worth subscale (α = .79), which is not included on the teacher-report version. Social Competence Scale (SCS). Teachers also completed the revised Teacher Social Competence Scale (SCS; Conduct Problems Prevention Research Group [CPPRG], 1995), a 17item measure assessing children’s prosocial/communication skills (7 items; e.g., “provide help, share materials, and act cooperatively with others”), emotion regulation skills (5 items; e.g., “handle disagreements in a positive way”), and academic skills (5 items; e.g., “performing academically at grade level”). Teachers rate each item using a six-point scale (0 = almost never, 1 = rarely, 2 = sometimes, 3 = often, 4 = very often, 5 = almost always). As recommended (CPPRG, 1995), the prosocial/communication and emotion regulation subscales were combined


11

in the present study to form a broader measure of social skills. In the present study, academic skills α = .92 and social skills α = .94 School Adjustment Questionnaire (SAQ) – Child Report. Children completed the School Adjustment Questionnaire (SAQ), a 20-item scale created for the Fast Track Project designed to evaluate children’s perceptions of their adjustment to school (Maumary-Gremaud, 2000). Completed at the end of the academic year in reference to overall functioning over the past year, children rate each SAQ item on a five-point scale (1 = never true, 2 = rarely true, 3 = sometimes true, 4 = usually true, 5 = always true). In the present study, the relationships with peers (6 items; e.g., “I had a hard time making friends at school this past year,” “other kids bothered me this past year”) and academic functioning/disciplinary actions (8 items; e.g., “this past school year was especially difficult for me,” “I had an easy time handling the new academic demands made on me”) scales were used, with some items reverse-coded before scoring. In the present study, relationships with peers α = .65 and academics α = .75. Loneliness Questionnaire (LQ). Children completed the Loneliness Questionnaire (LQ; Asher et al., 1984). Initially consisting of 24 items, a shortened 9-item version with superior psychometric properties (Ebesutani, Drescher et al., 2012) was used in the present study. These nine items (e.g., “It’s hard for me to make friends at school,” “I feel left out of things at school”) are rated on a three-point scale (0 = no, 1 = sometimes, 2 = yes). In the present study, α = .84. Children’s Emotion Management Scales (CEMS). Children completed the sadness (12 items) and anger (11 items) portions of the Children’s Emotion Management Scales (CEMS; Zeman et al., 2001). Each item is rated on a three-point scale (1 = hardly ever, 2 = sometimes, 3 = often). Factor analyses support 3 factors: inhibition (masking or suppressing emotional expression), dysregulated expression (culturally inappropriate emotional expression), and


12

emotion regulation coping (ability to appropriately cope with and control emotional experiences) (Zeman et al., 2001). Given our interest in examining children’s general emotion regulation, the sadness and anger items were combined to create composites of emotion inhibition (α = .79), dysregulation (α = .69), and coping (α = .81). Procedures All study procedures were approved by the Miami University Institutional Review Board (IRB). The principal investigator described the study to the eight teachers of grades three through six (all of whom were female). There were no exclusion criteria limiting teacher participation in this study. Teachers were told that study participation included completing measures for participating students and that they could withdraw their consent at any time. All eight eligible teachers (i.e., mainstream classroom teachers of students in grades three through six) provided signed informed consent to participate in the study. After teachers provided informed consent, the study was described by research staff to the students in each teacher’s classroom. Students were explicitly told that whether or not they participated in the study would have no effect on their grades. After answering any student questions, students were given informed consent forms for them to take home to their parents. The parent consent form informed parents that student participation in the study was fully optional, that providing consent was allowing for the student themselves and the student’s teacher to complete forms regarding their child, and that parents could revoke consent for participation in the study at any time. Parents were also given the e-mail address and phone number of the research team in the event that they had any questions or concerns. Students had two weeks to return the consent forms to the school. After one week, teachers were prompted by research staff to give students who had not yet turned in the parent informed consent form a new


13

copy to take home. Of the 199 total students in grades three through six at the time informed consent was obtained, 161 (81%) returned their consent forms. Of those, 131 (81% of those who returned their consent forms) provided consent for their child to participate in the study, and 126 (96%) were still attending the school when the data used in the current analyses were collected (as described below, two children were excluded due to validity concerns, resulting in a final sample size of 124). Teachers were given a packet with the study measures to complete in reference to each participating student and were asked to complete the packets within two weeks. Each teacher rated between 11 and 21 students (Median = 15 students). Teachers were compensated $7 for each packet they completed. Participating students from each grade completed the surveys in a group setting, using rooms in the school (e.g., cafeteria, gym) with adequate space to ensure students’ privacy. Prior to completing measures, children provided verbal assent. Specifically, students were told that their parents had given permission for them to answer questions for a project being conducted at their school but that their participation was fully optional. No student whose parent provided informed consent declined to participate in the study. Each child had their own packet of measures on which to record their responses. Research staff were continuously present to monitor pacing and answer questions. Student questions were addressed confidentially and individually, with children alerting staff of questions by raising their hand. Although no time limit was set for the completion of the measures, in a few cases ( .95, Tucker-Lewis Index (TLI) > .95, and root mean square error of approximation (RMSEA) < .08 (Hu & Bentler, 1999, Yu, 2002). A newer fit index, the weighted root square residual (WRMR), has been proposed for evaluating models using categorical indicators and weighted least squares estimation. Although less studied, simulation studies suggest that WRMR 10 are typically considered problematic) and all tolerance values were above .12 (values