Measures
Parenting stress. Parenting stress was assessed using the Parenting Stress Index 4th Edition – Short Form (PSI-4-SF) [40], an abbreviated version of the full-length Parenting Stress Index 4th Edition (PSI-4). The PSI-4-SF is measure of the overarching domains of parenting stress in which participants are asked to respond to 36-items using a 5-point Likert scale ranging from ‘strongly agree’ to ‘strongly disagree’. The PSI-4-SF is delineated into three subscales, including: parental distress, parent-child difficult interaction, and difficult child, with these three domains combined to form a total stress scale. Scores obtained within the 16th to 84th percentile are considered normal, whilst scores in the 85th to 89th percentile are consider high and those in the 90th percentile or higher are considered clinically significant. The PSI-4-SF has been found to show good internal consistency for its subscales, ranging from .88 (difficult child) to .95 (total stress), test-retest reliability and content and construct validity [40]. In the present analysis, total stress scale, parenting distress and parent-child dysfunctional interaction were of primary interest. Cronbach’s alpha estimates for the three subscales of the PSI-4-SF showed excellent reliability for parent-child difficult interaction (α = .91) and total stress score (α = .95), and good reliability for parental distress (α = .88), and difficult child (α = .89).
Parenting competence. Parenting competence was assessed using the Parenting Sense of Competence Scale (PSOC) [41]. The PSOC is a 17-item self-report measure with each item rated on a 6-point Likert scale ranging from 1 ‘strongly disagree’ to 6 ‘strongly sgree’. The PSOC is further differentiated into two subscales: parenting efficacy and parenting satisfaction, with higher scores indicating increased efficacy and satisfaction. The PSOC has been reported as having adequate reliability and validity in the literature [41, 42]. In the present analysis, Cronbach’s alpha estimates also showed good reliability parental satisfaction (α = .82) and acceptable reliability for parental efficacy (α = .77).
Personality traits and severity. Personality was assessed dimensionally across several different personality trait domains using the Personality Inventory for the Diagnostic and Statistical Manual of Mental Disorders 5th edition (DSM-5) – Brief Form (PID-5-BF) [2]. The PID-5-BF was developed by the DSM-5 Personality and Personality Disorders workgroup for the assessment of the alternative trait model for DSM-5. It is a 25-item self-report measure with each item rated on a 4-point Likert scale ranging from 0 ‘very false or often false’ to 3 ‘very true or often true’. The PID-5-BF assesses the overall severity of personality disorder, with scores ranging from zero to 75 and higher scores indicating greater overall personality dysfunction. The measure also assesses five personality trait domains including negative affect, detachment, antagonism, disinhibition, and psychoticism, with each trait domain ranging in score from zero to 15 and higher scores indicating greater dysfunction in the specific personality trait domain. The PID-5-BF is a relatively novel measure but has been found to have adequate internal consistency for personal trait domains and validity as a screening measure of dimensional maladaptive personality traits [43]. In our study, Cronbach’s alpha estimate for these five different personality trait domains were considered to demonstrate acceptable reliability for negative affect (α = .78), detachment (α = .71), disinhibition (α = .75) and psychoticism (α = .77), and questionable reliability for antagonism (α = .63). Personality pathology was also assessed categorically using the McLean Screening Instrument for Borderline Personality Disorder (MSI-BPD) [44]. The MSI-BPD is a 10-item binary measure of borderline personality symptomology, with a score of seven or more indicating that a person may meet criteria for a diagnosis of BPD. The MSI-BPD has shown good test-retest reliability [44], concurrent validity [45] and criterion validity [46]. In the current study, Cronbach’s alpha estimate for the MSI-BPD showed good reliability (α = .84).
Psychological wellbeing. Psychological wellbeing was assessed using the Mental Health Inventory (MHI-5) [47]. The MHI-5 is derived from the mental health dimension of the Medical Outcomes Study 36-Item Short Form Health Survey (SF-36). The MHI-5 is a 5-item self-report measure that participants respond to using a 6-point Likert scale ranging from 1 ‘all of the time’ to 6 ‘none of the time’. The MHI-5 has been found to be valid, and reliable for use with different subgroups and a range of different cultures [48]. In the current study, Cronbach’s alpha estimates for the MHI-5 showed good reliability (α = .88).
Trauma history. Parent’s own trauma history was assessed using the Childhood Trauma Questionnaire (CTQ-SF) [49]. The CTQ-SF is a 28-item self-report that screens for histories of abuse and neglect on a 5-point Likert scale ranging from 1 ‘never true’ to 5 ‘very often true’. The CTQ-SF enquires about five different types of maltreatment, namely emotional, physical, and sexual abuse, and emotional and physical neglect. The CTQ-SF has been validated with data from 2,000 clinical and non-referred participants by its authors [49]. Reliability coefficients for the CTQ-SF range from .92 for sexual abuse to .66 for physical neglect, test-retest intraclass correlations ranged from r = .79 for physical neglect to overall r = .86. Satisfactory content, construct, and concurrent validity was also reported. In the present study the CTQ-SF also showed excellent reliability for emotional abuse (α = .91), physical abuse (α = .90), and sexual abuse (α = .96), good reliability for physical neglect (α = .80), and questionable reliability for emotional neglect (α = .62).
Attachment style. Adult attachment style was assessed using the Relationship Questionnaire (RQ) [50]. The RQ is a 4-item self-report measure based on Bowlby’s [51] ‘inner working model’ in which participants are asked to respond to four paragraphs representing four attachment styles (secure, dismissing, preoccupied, fearful) using a 7-point Likert scale ranging from 1 ‘disagree strongly’ to 7 ‘agree strongly’. The RQ is founded on a sound theoretical basis and has good convergence with other attachment measures [52, 53] and has been shown to be predictive of therapy rates of change [54]. However, its reliability is less well studied because the nature of the scale does not lend itself to reliability estimates [55].
Reflective capacity. Participants’ parental reflective capacity of their ability to reflect on the mental states of their child and their role as a parent was assessed using the Parental Reflective Function Questionnaire (PRFQ) [56]. The PRFQ is a 36-item self-report measure that utilises a 7-point Likert scale ranging from 1 ‘strongly disagree’ to 7 ‘strongly agree’. The PRFQ is comprised of three subscales, namely: non-mentalizing mode, certainty about mental states and interest/curiosity in mental states. The PRFQ is a relatively novel measure, yet preliminary research suggests that it has good reliability and validity [56]. In this study the PRFQ showed acceptable reliability for non-mentalizing mode (α = .71), certainty about mental states (α = .76) and interest/curiosity about mental states (α = .72).