Correction: Maternal adverse childhood experiences and preschool children's behavioral problems: exploring mediation via an adapted measure of adult attachment pattern

Résumé fourni par la source

Correction on: Kurbatfinski S, Dosani A, Hayes AF, Dewey D and Letourneau N (2025) Maternal adverse childhood experiences and preschool children's behavioral problems: exploring mediation via an adapted measure of adult attachment pattern. Front. Psychol. 16:1596613. doi: 10.3389/fpsyg.2025.1596613 Adding/removing textA correction has been made to the section [Methods, Included Variables and Measures, Mediator Variable, Pages 04-05]: Original Writing: "Mothers' adult attachment pattern, rated from secure (low score) to insecure (high score), as well as degree of insecure attachment pattern subtypes (dismissive and preoccupied), were coded from the Revised Experiences in Close Relationships Questionnaire (ECR-R) (Fraley et al., 2000). This was completed when children were about 60 months of age. The ECR-R is a self-report measure that includes 12 questions rated on a Likert Scale ranging from 1 (strongly disagree) to 7 (strongly agree). Since the ECR-R shows similar psychometric properties compared to the original 36-item ECR (Brenning et al., 2014;Brugnera et al., 2019), we chose to use the shortened version to limit participant burden. Of the 12 questions, 6 are relevant to preoccupied and 6 to dismissive adult attachment patterns, respectively. Mothers' adult attachment pattern was calculated by summing the preoccupied and dismissive adult attachment pattern scores deriving from the 12 questions. Scores for mothers' adult attachment pattern therefore ranged from 12 to 84 with lower scores indicating a more secure adult attachment pattern while higher scores indicated a more insecure adult attachment pattern. Scores for the preoccupied and dismissive adult attachment pattern subtypes were derived from their respective 6 questions and ranged from 6 to 42 (Fraley et al., 2000). The ECR-R demonstrates high validity and reliability in assessing for adults' dismissive and preoccupied adult attachment patterns (Sibley et al., 2005). While it has not yet been used in research to assess for an overall adult attachment pattern (secure versus insecure), Fraley, the developer of the ECR-R, argues that insecure adult attachment pattern is an additive combination of both preoccupied and dismissive adult attachment pattern behaviors (Fraley and Waller, 1998). We calculated a McDonald's Omega coefficient of 0.87 across all 12 items producing the continuous score of adult attachment pattern, suggesting high internal reliability consistency. Additionally, there was a strong general factor driving most items and two group factors contributing additional variance (although one was much stronger than the other). These findings suggest that while the measure is bidimensional, the total score reasonably measures adult attachment pattern continuously from secure to insecure and is appropriate to use in analysis. Overall, this serves as a foundational study employing the ECR-R to measure mothers' adult attachment pattern as a continuum of secure to insecure."Corrected paragraph: "Mothers' adult attachment pattern, rated from secure (low score) to insecure (high score), as well as degree of insecure attachment pattern subtypes (dismissive and preoccupied), were coded from the Experiences in Close Relationships-Short Form (ECR-S) (Wei et al., 2007). This was completed when children were about 60 months of age. The ECR-S is a self-report measure that includes 12 questions rated on a Likert Scale ranging from 1 (strongly disagree) to 7 (strongly agree). Since the ECR-S shows similar psychometric properties compared to the original 36-item ECR measure (Brennan et al., 1998;Brenning et al., 2014;Brugnera et al., 2019), we chose to use the shortened version to limit participant burden. Of the 12 questions, 6 are relevant to preoccupied and 6 to dismissive adult attachment patterns, respectively. Mothers' adult attachment pattern was calculated by summing the preoccupied and dismissive adult attachment pattern scores deriving from the 12 questions. Scores for mothers' adult attachment pattern therefore ranged from 12 to 84 with lower scores indicating a more secure adult attachment pattern while higher scores indicated a more insecure adult attachment pattern. Scores for the preoccupied and dismissive adult attachment pattern subtypes were derived from their respective 6 questions and ranged from 6 to 42 (Wei et al., 2007).The ECR-S demonstrates high validity and reliability in assessing adults' dismissive and preoccupied adult attachment patterns (Wei et al., 2007). While it has not yet been used in research to assess an overall adult attachment pattern (secure versus insecure), prior work suggests that insecure adult attachment reflects an additive combination of both preoccupied and dismissive adult attachment pattern behaviors (Fraley & Waller, 1998). We calculated a McDonald's Omega coefficient of 0.87 across all 12 items producing the continuous score of adult attachment pattern, suggesting high internal reliability consistency. Additionally, there was a strong general factor driving most items and two group factors contributing additional variance (although one was much stronger than the other). These findings suggest that while the measure is bidimensional, the total score reasonably measures adult attachment pattern continuously from secure to insecure and is appropriate to use in analysis. Overall, this serves as a foundational study employing the ECR-S to measure mothers' adult attachment pattern as a continuum of secure to insecure."A correction has been made to the section [Methods, Missing Data, Page 05]: Original Writing: "Although 1,236 participants provided ACEs data at 1 year of age, only 636 provided data on child behavioral problems at 5 years of age (a 48% attrition rate over 4 years). Some variables, such as mothers' income and depressive symptoms, had missing data that constituted less than 5% of the overall sample; however, missing data for mothers' anxiety symptoms constituted more than 5% of the sample (n = 48). We assessed missing data patterns for mothers' anxiety via two methods to determine best methods in handling the missing data. Results from a binary logistic regression with maternal anxiety as a binary outcome (zero indicated data present, one indicated missing data) revealed that only maternal age predicted missingness of anxiety symptom data (p < 0.05). However, Little's Test yielded a significant test statistic (p < 0.05), suggesting data were not missing completely at random. Taken together, we deemed the missing data for mothers' anxiety symptoms as missing at random. Therefore, all missing data were imputed via multivariate imputation by chained equations using an average of ten imputations, a commonly employed method for addressing data missing at random (Austin et al., 2021)."Corrected paragraph: "Although 1,236 participants provided ACEs data at 1 year of age, only 636 provided data on child behavioral problems at 5 years of age (a 48% attrition rate over 4 years). Some variables, such as mothers' income and depressive symptoms, had missing data that constituted less than 5% of the overall sample; however, missing data for mothers' anxiety symptoms constituted more than 5% of the sample (n = 48). We assessed missing data patterns for mothers' anxiety via two methods to determine appropriate approaches for handling missingness. Results from a binary logistic regression with maternal anxiety as a binary outcome (0 = data present, 1 = missing) revealed that only maternal age predicted missingness of anxiety symptom data (p < 0.05). However, Little's Test yielded a significant test statistic (p < 0.05), suggesting data were not missing completely at random. Taken together, the missing data for mothers' anxiety symptoms were deemed consistent with a missing at random (MAR) assumption. Missing data were addressed using multivariate imputation by chained equations (MICE) with twenty imputations, a commonly employed method for addressing data

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.