Development and validation of a shortened version of the Child Abuse Self Report Scale (CASRS-12) in the Arabic language

Child maltreatment refers to “all forms of physical and emotional ill-treatment, sexual abuse, neglect, and exploitation that results in actual or potential harm to the child’s health, development or dignity” [1, 2]. Four main types of child abuse can be identified: physical, sexual, and psychological abuse (acts of commission), and neglect (act of omission in the care) [3]. A systematic review and meta-analysis estimated that the global number of children aged between 2 and 17 years old who were victims of any form of abuse (physical, emotional and sexual) was of over 1 billion [3]. There is a strong evidence that child abuse contributes to short- and long-term detrimental consequences on physical and mental health [4‐7]. Beyond the immediate pain and hurt it engenders, child abuse causes a wide range of negative effects, including bodily pain poor general health, depression, anxiety [8], suicide ideation [9], posttraumatic stress symptoms, dissociation, aggression, social withdrawal, school absences, suspension, or withdrawal [10], life-lasting cognitive deficits [11], bullying victimization [12, 13], violence perpetration and crime [14]. Child abuse was referred to as the “invisible epidemic” [15], since it has affected and continues to affect a substantial proportion of children worldwide [16]. As such, child abuse has been identified by the World Health Organization as a major risk factor related to the global burden of disease [17]. It has also been recognized since decades and globally as major, but preventable, public health and human rights problem [18]. The problem is more critical in some contexts such as low- and middle-income countries [19], more particularly Arab countries [20].

Arab countries are demographically one of the most youthful countries in the world [21]. Physical and psychological violence against children as a form of discipline is largely normalized and accepted in Arab cultures, and is not legally prohibited in the vast majority of Arab countries [22]. Due to the many conflicts in the region, the new Arab generations have known a dramatic increase in collective violence during the last years [22]. The UNICEF estimated that, in 2015, more than 70% of the world’s adolescents who died due to collective violence live in the Middle East and North Africa, and 7 in 10 children living in the region are physically abused [22]. Therefore, efforts to effectively assess, monitor, and manage the consequences of child abuse and neglect on Arab adolescents and young adults should be prioritized in the region. In addition, and given the magnitude of the problem, academic institutions have been called upon to increase research capacity to build a local evidence-base on violence against children in Arab countries that helps inform policies and interventions [23]. The first step to this end is to provide valid measurement instruments that can be used in Arab settings to evaluate the multidimensional and complex construct of child abuse.

A few scales have been previously used in Arab studies, including the 30-item Adverse Childhood Experiences-International Questionnaire (ACE-IQ) [24] (e.g., in Saudi Arabia [25], Iraq [26], Tunisia [27]), the 28-item Childhood Trauma Questionnaire (CTQ) [28] (e.g., in Tunisia [29, 30], Saudi Arabia [31], Egypt [32]), and the Child Abuse Self Report Scale (CASRS) [33] (e.g., in Lebanon [34]). The CASRS consists of 38 items that load onto four subscales assessing four abuse dimensions: (1) Physical abuse (physical punishment/abuse), (2) Psychological abuse (acts leading to fear or psychological pain), Sexual abuse (unwanted sexual touch and forced sexual contact by an adult or older child, including family members), and Neglect (lack of engaging in behaviours that are necessary to meet the developmental needs of a child, e.g. failure to provide adequate supervision or food) [33]. All these dimensions are accounted for, regardless of whether there was injury or damage caused [33]. The CASRS has previously been translated by our team using the forward and backward method [34], and has exhibited excellent psychometric properties in both Arabic-speaking clinical [35] and non-clinical populations [34, 36]. However, in the specific context of abuse and trauma, all these scales may be challenging to administer due to their length. There have been a very few validation studies of short child maltreatment measures (e.g., [37, 38]); however, no brief forms exist in Arabic so far. A higher number of items is associated with more administration time and less motivation to engage in responding. Longer scales are thus linked to lower quality responses, higher refusals and drop-outs, and lower response rates lower [39]; in addition to being costly. One strategy to overcome these disadvantages is to develop briefer forms that offer the potential benefits of being more practical, easier to interpret, less time-consuming and burdening, less costly; while trying at the same time to preserve the same measurement precision and standards of psychometric excellence of the full-length version [40]. In this perspective, we sought to develop a brief version of the Arabic CASRS that would assess the construct of child abuse in an appropriate and valid way. We thus aimed to explore its factorial structure, internal consistency, divergent validity, and correlations with other measures. We hypothesized that the Arabic CASRS-12 would (1) reproduce the four-factor structure identified by the developers of the original 38-item CASRS, (2) show adequate validity and reliability, (3) be invariant across gender.

Arab countries have levels of child abuse that are among the highest globally. We believe that providing an Arabic brief measure of child abuse that permits to decrease respondent burden and costs of data collection while preserving data quality may be potentially helpful for clinicians and highly useful for researchers and policy makers in the developing Arab countries. We thus aimed through the present study to develop and validate a brief and psychometrically sound version of the CASRS, as a reliable and valid alternative to the already existing 38-item version that has been widely used in the Lebanese context. As expected, we found good model fit for the four-factor solution, adequate composite reliability, good divergent validity, as well as configural, metric, and scalar invariance across gender. One potential strength of this scale is that it assesses the four universally consensual dimensions of the child maltreatment construct (i.e., Physical abuse, Psychological abuse, Sexual abuse, Neglect) [3] regardless of whether there was injury [1, 2] through only 12 items. In light of these findings, we potentially encourage clinicians and researchers to use this scale as a valid and reliable measure of child abuse among Arabic-speaking populations.

While there have been a range of measures to assess child maltreatment, their cross-cultural validity is still largely unknown; especially in certain contexts where data is yet scarce [63]. However, many aspects of childhood trauma are largely influenced by culture. Cultural norms and values normalize to some extent some forms of abuse/neglect in some contexts. In the Lebanese society, for example, it is “normal” for children to self-care without their parents’ supervision at ages younger than what is commonly accepted [64]. Additionally, violence perpetuated within the family system can be seen to be not harmful and even adequate parental supervision in collectivist societies [64]. For these reasons, we chose to validate a brief form of a scale that was developed in a Middle East country and a collectivistic society, which might be more suitable for Arab people than all other scales that were mostly developed in Western and individualist countries.

Results revealed that both EFA and CFA yielded a four-factor structure for the CASRS-12 that mirrors the original four factors captured by the original CASRS [33], further supporting the multidimensional factor structure interpretation of the scale. In addition, the Arabic CASRS-12 showed a good internal consistency as evidenced through McDonald’s ω values ranging from 0.87 to 0.93 for the four subscales. This is consistent with the original validation study where a strong internal consistency was attested by Cronbach alpha values ranging from 0.82 to 0.95 [33], and other previous studies using the CASRS in various contexts and settings [34‐36]. We consider that using McDonald’s ω strengthens our findings since it has several advantageous over Cronbach’s alpha when assessing the internal consistency of multidimensional measures [65].

In addition, our results indicate evidence for measurement invariance across gender, proving that the CASRS-12 can be applied to make valid comparisons between male and female respondents. In this vein, we found that our male participants reported having experienced significantly more physical abuse, sexual abuse and neglect compared to females. In agreement with our findings, multiple studies in different Arab countries (e.g., Palestine [66], Egypt [67], Lebanon [68]) have shown that all forms of violence and abuse are more prevalent in men than women.

In order to attest for divergent validity of the scale, and based on previous literature, we examined the correlations between child abuse dimensions and bullying victimization, eating attitudes and social support. We found that all CASRS-12 dimensions were significantly and positively correlated with more inappropriate eating attitudes and lower social support; and that psychological, physical, and sexual abuse correlated with higher bullying victimization. These findings confirm discriminant validity of the scale; and are in line with previous literature stipulating that child abuse is closely related to a range of mental health and behavioral problems including bullying victimization [12, 13], eating disorders [69, 70], and lower levels of perceived social support in adulthood [71, 72]. However, to further confirm the clinical utility of the Arabic CASRS-12, additional validation studies in clinical populations are required.

Our study has certain limitations that need to be discussed. First, we used a cross-sectional design that limits conclusions about causality. Second, the study adopted an online questionnaire which, while being more effective in preserving anonymity and reducing social desirability, may be a source of limited generalizability and selection bias. Third, we used a population-based sample that precludes any conclusions about the validity of the scale in clinical samples. Fourth, further research need to investigate the stability of the CASRS-12 through test–retest reliability.

Our findings support the reliability and validity of the Arabic four-factor structure CASRS-12. We hope that providing this simple and easy-to-use short form of the scale will broaden its utilization across different settings among Arabic-speaking people, and encourage cross-cultural research on child abuse, and help inform local prevention and intervention strategies.