Author Queries
Total Page:16
File Type:pdf, Size:1020Kb
Author Queries Journal title: ASMNT Article Number: 10.1177/1073191114568301 Dear Author/Editor, Greetings, and thank you for publishing with SAGE. Your article has been copyedited, and we have a few queries for you. Please respond to these queries when you submit your changes to the Production Editor. Thank you for your time and effort. NOTE: Please answer the queries by editing within the article: follow the AQ hyperlink below and edit the text directly. Add a "New Comment" (in the "Review" tab), if the correction cannot be done directly or for further queries. Below, please tick the boxes to the right of the query, to indicate the query has been dealt with. Please assist us by clarifying the following queries: AQ 1 Please confirm that all author information, including names, affiliations, sequence, and contact details, is correct. X Eliminato: Please review the entire document for typographical errors, mathematical errors, and any other necessary corrections; check X AQ 2 Eliminato: headings, tables, and figures. Please confirm you have reviewed this proof to your satisfaction and understand this is your final opportunity for review prior to X Eliminato: AQ 3 publication. AQ 4 Please confirm that the Funding and Conflict of Interest statements are accurate. X Eliminato: Article Assessment 1–12 Evaluating Measurement Invariance © The Author(s) 2015 Reprints and permissions: Between Parents Using the Strengths and sagepub.com/journalsPermissions.nav DOI: 10.1177/1073191114568301 Difficulties Questionnaire (SDQ) asm.sagepub.com Carlo Chiorri 1, James Hall 2, Jeffrey Casely-Hayford 3, and Lars-Erik Malmberg 2[AQ1][AQ2][AQ3] Abstract Parent ratings of their children’s behavioral and emotional difficulties are commonly collected via the Strength and Difficulties Questionnaire (SDQ). For the first time, this study addressed the issue of interparent agreement using a measurement invariance approach. Data from 695 English couples (mothers and fathers) who had rated the behavior of their 4.25-year-old child were used. Given the inconsistency of previous results about the SDQ factor structure, alternative measurement models were tested. A five-factor Exploratory Structural Equation Model allowing for nonzero cross-loadings fitted data best. Subsequent invariance analyses revealed that the SDQ factor structure is adequately invariant across parents, with interrater correlations ranging from .67 to .78. Fathers reported significantly higher levels of child conduct problems, hyperactivity, and emotional symptoms, and lower levels of prosocial behavior. This suggests that mothers and fathers each provide unique information across a range of their child’s behavioral and emotional problems. Keywords measurement invariance, exploratory structural equation modeling, multitrait–multisource model, Strengths and Difficulty Questionnaire The Strengths and Difficulties Questionnaire (SDQ; available). Thus, interparent agreement is a crucial issue Goodman, 1994) is a user-friendly instrument designed for for all contexts in which the SDQ is used and where both the assessment of behavioral and emotional problems in parents are informants. If it can be shown that both parents children and adolescents aged 3 to 16 years. Its consists of provide the same information, then either can be 25 items equally divided across five scales, four of which confidently used as a single informant. On the other hand, probe difficulties: emotional symptoms (EMO), conduct if systematic differences exist between parents, then the problems (COND), hyperactivity–inattention (HYP), and use of both parental reports would be advisable to enhance peer problems (PEER); and one scale which probes the sensitivity of scores in identifying children requiring strengths: prosocial behavior (PRO). The SDQ adopts a clinical attention, as they add relevant unique variance multi-informant approach, with versions designed for (e.g., De Los Reyes & Kazdin, 2005). teachers and parents, and a self-report version for One common and potentially problematic feature of adolescents aged 11 to 16 years. Given its brief format, it studies on interparent agreement is that they usually rely is widely used as a screening tool in research, education, on assumptions that might not be supported by empirical community, and clinical settings. evidence. Analyses are usually carried out assuming that With translations into over 60 languages (Stone, Otten, there is measurement invariance of the measures between Engels, Vermulst, & Janssens, 2010) there has been much mothers and fathers. To the best of our knowledge, no evaluative research on most of the psychometric properties study has addressed this issue on the SDQ. This is of the SDQ. Factor structure, internal consistency, important because to compare scores across parents, it construct (convergent and discriminant) validity, and criterion (concurrent and predictive) validity have been 1 extensively tested, but very few studies have investigated University of Genoa, Genoa, Italy 2University of Oxford, Oxford, UK the interparent agreement of SDQ scores relating to their 3University of Bath, Bath, UK child. This is somewhat surprising, as in most studies a not-otherwise-specified “parent” provided the data, Corresponding Author: without reporting whether it was the mother or the father Carlo Chiorri , Department of Educational Sciences, University of Genoa, (let alone those cases in which only one parent is Corso A. Podestà, 2, Genoa 16128, Italy. Email: [email protected] 2 Assessment must be shown that the latent dimensions that underlie the and .68; Fathers: .69 and .68; PEER: Mothers .32 and .29; SDQ measure the same construct in the same way, and that Fathers: .29 and .25; PRO: Mothers .66 and .60; Fathers: .67 the measurements themselves operate in the same way and .61) other than for paternal ratings of boy’s COND across parents. Otherwise, mean differences and other (Mothers .42 and .56; Fathers: .44 and .40). Further analyses comparisons are likely to be invalid. The purpose of this were then carried out after the scores on COND and HYP study is thus to fill this gap in the literature. were combined into a single externalizing problems score, and excluding PEER due to its low reliability. Pearson correlations between mother and father scores were in the Interparent Agreement on Children’s Emotional and moderate to strong range: .61 and .59 for girls and boys, Behavioral Problems respectively (.40 and .46 for EMO and .46 and .38 for Past research on interparent agreement via reports of PRO). Furthermore, mother and father reports of the emotional and behavioral problems in children and behavioral difficulties of their daughters agreed, whereas adolescents has provided mixed results. For example, parents differed when rating the PRO of their sons (mothers while a meta-analysis by Achenbach, McConaughy, and reported significantly higher levels of PRO). Howell (1987) reported moderate, although significant, If we adopted the ratings suggested by Cicchetti and interparent agreement for both internalizing and Sparrow (1981; “poor” when lower than .40, “fair” when externalizing problems, a similar study by Duhig, Renk, ranging from .40 to .59, “good” when ranging from .60 to Epstein, and Phares (2000) found a moderate correlation .74, and “excellent” when higher than .7 5) then the Eliminato: 4 between mother and father ratings of internalizing interparent agreement coefficients reported by Davé et al. problems, but higher interparent correlations for (2008) and Mellor et al. (2011) would be considered Eliminato: and externalizing problems. Furthermore, interparent “poor” to “fair.” Thus, the results of these two studies agreement varied with the age of the children. Achenbach suggest only moderate agreement between parents on SDQ et al. (1987) found a higher agreement for younger scores and this is consistent not only with previous children, whereas Duhig et al. (2000) found higher research using other childhood measures of behavioral and agreement for adolescents. In contrast, one consistent emotional problems but also with studies on other traits finding within the interparent agreement literature is that such as anxiety (Moreno, Silverman, Saavedra, & Phares, mothers consistently report more behavioral and emotional 2008), psychopathology (De Los Reyes & Kazdin, 2005), problems than do fathers, although this may also depend Big Five personality (Tackett, 2011), and conflict and on the measure employed and the age of the child (for closeness in parent–child relationships (Driscoll & Pianta, reviews, see Davé, Nazareth, Senior, & Sherr, 2008; 2011). These common inconsistencies are thought to Mellor, Wong, & Xu, 2011). reflect real differences in the way that a child is perceived To the best of our knowledge, only two studies have by their parents and might stem from the tendency for investigated the consistency and differences between mothers and fathers to play different parenting roles and to mother and father ratings on the SDQ. Davé et al. (2008) engage in different activities with a child. For example, collected data from 248 dyads composed of British mothers might have a higher participation in childrearing biological mothers and fathers who were both residing activities and spend more time with a child, particularly with their own 4- to 6-year-old child. Cronbach’s alphas with infants and toddlers. Moreover, a child might behave were similar across all SDQ scales (PRO: .69 vs. .70 for differently when alone with one of the parents