Confirmatory factor analysis of the Maltreatment and Abuse Chronology of Exposure (MACE) scale: Evidence for essential unidimensionality

Abstract

The Maltreatment and Abuse Chronology of Exposure (MACE) scale is a retrospective self-report scale that measures 10 distinct types of childhood maltreatment. Despite its increasing use, the factor structure of the MACE has not been thoroughly investigated. As such, the viability of the MACE total and subscale scores are uncertain. The current study investigated the factor structure of the MACE in order to quantify the reliability of its total and subscale scores. Two independent samples of participants (N=1051 & N=582) who completed the MACE were included in this study. Using confirmatory item response models, we tested one-factor and several bifactor models of participants’ responses. We used model-based indices to estimate the reliability of the MACE total and subscale scores, and to quantify the essential unidimensionality of the scale. We found the MACE total score was a reliable and valid measure of overall childhood maltreatment. In contrast, although we found MACE subscale scores exhibited adequate reliability, the vast majority of their reliable variance reflected general maltreatment and not any particular type of maltreatment as intended. We also found that the MACE is essentially unidimensional and some of its items exhibit differential item functioning by gender. Our results provide support for a one-factor structure of the MACE, as well as continued use of the MACE total score. Our results caution against the use of MACE subscale scores, which have little practical use insofar that they provide little unique, reliable information above and beyond the total score.

How adverse childhood experiences get under the skin: A systematic review, integration and methodological discussion on threat and reward learning mechanisms

“…Consider (additional) assessment tools which allow for a fine-grained evaluation of potentially relevant ACE characteristics including onset and duration of exposure and controllability (e.g. MACE, Teicher and Parigger, 2015 ; although following Zorowitz and Tuominen, 2022 it lacks subscale specificity) as well as social aspects (e.g. social support).…”

Watching with Argus eyes: Characterization of emotional and physiological responding in adults exposed to childhood maltreatment and/or recent adversity

“…As the MACE provides a more fine-grained characterization of early life adversity as well as its developmental timing and duration, we suggest favoring more detailed measurement instruments over less detailed instruments and focus in future work also on the instruments psychometric properties (Flake & Fried, 2020). Note, however, that the overall score of the Maltreatment and Abuse Chronology of Exposure (MACE; the English version of the KERF) has been shown to be reliable while the subscales lacked specificity (Zorowitz & Tuominen, 2022).…”

“…As the MACE provides a more fine‐grained characterization of early life adversity as well as its developmental timing and duration, we suggest favoring more detailed measurement instruments over less detailed instruments and focus in future work also on the instruments psychometric properties (Flake & Fried, 2020). Note, however, that the overall score of the Maltreatment and Abuse Chronology of Exposure (MACE; the English version of the KERF) has been shown to be reliable while the subscales lacked specificity (Zorowitz & Tuominen, 2022). In sum, we suggest that a multiverse‐type of approach (Lonsdorf et al, 2022; Simonsohn et al, 2020; Sjouwerman et al, 2021; Steegen et al, 2016) may shed light on the question to what degree different cutoffs (e.g., Bernstein et al, 2003; Häuser et al, 2011) for the classification into exposed/non‐exposed (despite the same questionnaire and sample) converge and to what extent different questionnaires, interviews and other instruments converge in their classifications or dimensional measures for childhood maltreatment in the same sample.…”