Adult recall of childhood psychological maltreatment: Definitional strategies and challenges

Abstract

This review paper was undertaken to compare legal definitions, conceptual definitions, and existing survey instruments of adult recall of childhood psychological maltreatment and to identify definitional challenges that remain. Fifteen measures were compared in order to determine whether the items on the measures captured all of the core constituent components of psychological abuse. Five key issues were identified for the field including, (1) lack of consistency in the term itself, (2) lack of consistency between legal and conceptual definitions, (3) lack of consistency between conceptual and operational definitions, (4) lack of clarity as to the role of witnessing domestic violence, and (5) lack of consistency in psychometric characteristics of the measures. Taken together these issues interfere with the ability of researchers to integrate findings across studies and to move the field forward. Suggestions for systematically addressing these issues are offered.

Introduction

The problem of psychological maltreatment of children is now well recognized. The extant literature is extensive and contains two distinct branches: (1) theory and research regarding adult recall of childhood psychological maltreatment with an emphasis on documenting its long-term effects and (2) theory and research regarding current psychological maltreatment of children with an emphasis on identification, prevention, and intervention. The purpose of this paper is to identify the current state of definitional strategies and challenges as they relate to the study of adult recall of childhood psychological maltreatment.

Development and testing of measures of retrospective adult recall of childhood psychological maltreatment has been a fruitful area of research for the past few decades with several measures having been developed and field-tested to one degree or another. The advantages of conducting research in this area – as opposed to conducting research on currently maltreating parents and their children – are many. To begin with, the ability to examine the long-term effects of various forms of psychological maltreatment on a range of outcomes is possible without having to wait several years, which would be necessary in a longitudinal design and without incurring the expense involved in tracking subjects and collecting data over multiple data collection sessions. A second benefit is that informed consent is less difficult to obtain for studies of adults than for children and their parents, which require more stringent informed consent procedures (that may also dampen participation). It is considerably easier to obtain Institutional Review Board (IRB) approval for research with community samples than with any kind of vulnerable populations. Thus, logistically and financially, adult retrospective studies with community samples is a less problematic program of research to plan and conduct than a prospective longitudinal study of children and families, especially those who are deemed to be at risk or vulnerable (Baker, 2008). In addition, studies employing adult recall methods allow the researcher to benefit from the wisdom and adult perspective of child victims of abuse. Only adults can reflect on the long-term impact of the abuse or provide an integrated assessment of the meaning and impact of it.

At the same time, there are inherent limitations to the measurement of adult recollections of any childhood events, especially ones that are likely to be reconstructed and processed over time such as abuse. Widom, Raphael, and DuMont (2004) provide a review of the potential pitfalls of studies that aim to establish consequences of abuse through simultaneously assessing retrospective reports of abuse and current reports of functioning, notably the likelihood that individuals whose functioning is compromised (i.e., are currently depressed, anxious, or dissatisfied) are likely to recall greater childhood difficulty either through a desire to explain and understand current circumstances or through distorted cognitive styles in which everything is perceived as more bleak and troubled than it really was. Several longitudinal studies have shed light on the extent of discrepancies of recall of childhood abuse, also suggesting that caution is required (e.g., White, Widom, & Chen, 2007). Nonetheless, there are good reasons to proceed with this line of inquiry, especially as it relates to the two primary purposes of assessing adult recollection of childhood abuse: (1) establish consequences of childhood abuse over the life span and (2) assess individuals at the time they are already adults, such as parents of children in the child welfare system.

Some adults come to the attention of researchers or clinicians at the time that they are already adults and it is not possible to measure child abuse with anything other than retrospective accounts. Such assessments (despite potential recall error and bias) may be fruitful as they reflect the individual's own experience and perception, which itself is meaningful even if not wholly reliable or accurate. For example, currently maltreating parents could be assessed for childhood abuse as part of an effort to understand and address their own unmet needs for safety and nurturance, a likely prerequisite to improving their own parenting given the intergenerational transmission of attachment security (e.g., Main, Kaplan, & Cassidy, 1985). In fact, there is good reason to expect that service agencies aiming to prevent reoccurrence of abuse in maltreating families would need some mechanism for assessing the parents' own experiences with all forms of abuse, including psychological. History of abuse is clearly a potentiating risk factor, increasing the probability of the occurrence of maltreatment (Cicchetti & Rizley, 1981).

For these reasons, adult retrospective studies of childhood maltreatment represent an important area of clinical need and research attention and have proven to be a fruitful area of empirical investigation. However, key definitional issues remain unresolved – including varying and inconsistent legal, conceptual and operational definitions – which continue to interfere with the ability of researchers and practitioners to fully integrate and apply the data that have been collected.

To begin with, there is no uniform legal definition of each type of child abuse, including psychological. Table 1 presents a summary of state definitions (Child Welfare Information Gateway, 2008).

As can be seen, most of the legal definitions of psychological maltreatment refer to the impact on the child, not on the parental acts that may have led to such a result. Even within this child outcome focus, there is considerable variation across states. Several definitions utilize the phrase, “observable and sustainable impairment in the child's ability to function in a developmentally appropriate manner” or comparable language and a few elaborate on the behavioral signs that a child has been harmed referring to severe anxiety, depression, withdrawing, and outwardly aggressive behavior; and only a handful describe the specific behaviors that may cause such outcomes other than general terms such as “acts or omissions” or “emotional abuse.” About one fifth of the state definitions describe the person inflicting the emotional abuse, usually the parent, guardian, caretaker, custodian, or person responsible for the child. Thus, legal definitions apply a child outcome approach and define it as something that results in impairment to the child's functioning rather than focus on the parental behaviors that would result in such outcomes. While states are consistent in this respect, there is still variation regarding the inclusiveness of the law (how broad or loose the definition is) (Hamarman, Pope, & Cjaja, 2002). Likely due to variations in inclusiveness, rates of reported emotional abuse have been found to vary widely across states from .37 cases in 10,000 (Pennsylvania) to 113.02 cases per 10,000 (Connecticut), (Hamarman et al., 2002).

While legal definitions utilize a child outcome approach, conceptual definitions reflect greater variation, with some employing a child outcome focus and others employing a parental standard of care approach. However, even within type, (child outcome or parental behavior) there is considerable variation. Table 2 presents a summary of the key conceptual definitions that employ a child outcome approach.

Three different types of child outcomes appear to be relevant for defining psychological maltreatment. The first defines the construct in general terms as something that is psychologically harmful or damaging to the child (e.g., Faller et al., 1981, Garbarino and Vondra, 1987, Hart et al., 1983a, McGee and Wolfe, 1991). The second type defines it as something that harms the child's development in one or more functional areas such as the social, cognitive, affective, behavioral, or emotional domain (e.g., Hart et al., 1987, Moran et al., 2002). The third type defines psychological maltreatment as something that affects the child's perception of him- or herself as good and the world as safe, and employs concepts such as self-esteem, competence, achievement, belonging, and safety (e.g., Garbarino, 1978, Garbarino and Gilliam, 1980, Iwaniec, 1995). What these definitions share is a belief that the specific parental behaviors that result in these outcomes are either irrelevant or too many to enumerate and that what matters most is the effect on the child. This approach allows for the fact that different children at a single point in time or the same child at different points in time may be adversely affected by different parental behaviors. What results in psychological maltreatment varies depending upon the circumstances and the individual child.

Table 3 presents a summary of the major conceptual definitions that employ the alternative approach, focusing on parental behaviors that are likely to result in psychological harm to the child.

As can be seen, there are many different types of behaviors that can be considered psychologically maltreating. There is some overlap in the specific parental behaviors considered reflective of psychological maltreatment, although there are also many behaviors mentioned in only one definition. The following behaviors are included in at least one definition: Berating/disparaging, cruelty, denying emotional responsiveness/ignoring, developmentally inappropriate or inconsistent interactions with the child, developmentally inappropriate expectations of child, exploiting/corrupting, failing to recognize or acknowledge the child's individuality and psychological boundary, failing to promote the child's social adaptation, hostility, inadequate nurturance/affection, inappropriate emotional responses to the child's emotional expressions, isolating/close confinement, mental health/medical/educational neglect, mis-socializing, negative attributions and misattributions to the child, over-pressuring, rejecting, spurning, terrorizing, threatening harm, verbally assaulting, punishing an infant's operant social behavior, and punishing a child's manifestations of self-esteem. Most definitions include about five behaviors with the range being one to seven.

With only minimal fine-tuning, all of these behaviors can be classified into one of the five categories of psychological maltreatment identified by the American Professional Society on the Abuse of Children APSAC (Binggeli, Hart, & Brassard, 2001) which was drawn from the pioneering work of Garbarino, Guttman, and Seeley (1986): (1) spurning (being mean, berating, disparaging, screaming and yelling at, verbally assaulting, cursing and swearing at, calling derogatory names, being cruel, teasing, being hostile, rejecting, negatively comparing child to others, treating child like black sheep); (2) terrorizing (threatening harm to child or others, threatening to abandon); (3) isolating the child from friends and family and normal socialization experiences, close confinement; (4) exploiting/corrupting (being or encouraging the child to be amoral/corrupt, mis-socializing, encouraging developmentally inappropriate behavior such as infantilizing or parentification, encouraging or coercing abandonment of developmentally appropriate autonomy through over involvement and intrusiveness), and (5) denying emotional responsiveness (being cold or indifferent, providing inadequate nurturance/affection, ignoring, being detached).¹ Although the APSAC (Binggeli, Hart, & Brassard, 1981) definition appears to represent a useful tool for synthesizing other existing parent-focused definitions, such a consensus has not yet occurred among researchers in the field. Thus, even if there were consensus that the construct should be defined by parental behaviors, there is still no definitive agreed upon list of what those behaviors would be.

Further, a recent review of the literature revealed mixed support for the negative impact of these various behaviors on the child, which should be the primary determination as a form of abuse (Brassard & Donovan, 2006). In their review of the body of evidence as it relates to the APSAC definitions, Brassard and Donovan conclude that the evidence regarding spurning and terrorizing is quite strong (e.g., Maughan et al., 1995, Wolfe and McGee, 1994). Denying emotional responsiveness, especially early in life when the parent is the sole source of stimulation, has been strongly confirmed as a form of psychological maltreatment (England & Erikson, 1987). Evidence of the effects of isolating was deemed less compelling, while evidence for corrupting/exploiting was found to be lacking as the area is relatively understudied. The authors conclude that more research is required before there is sufficient empirical support for the five categories in the APSAC definitional schema.

In light of the definitional variation at the legal and conceptual levels, it is not surprising that operationally there is variation as well. Table 4 presents a summary of self-report survey measures that have been developed to assess adult recall of childhood psychological maltreatment. These measures were identified primarily through an extensive review of the literature based on numerous searches of PsychInfo.

Fifteen scales have been created to assess psychological maltreatment. Each is briefly described below. The Child Abuse Questionnaire (CAQ) was first reported on in Gross and Keller (1992), although the authors attribute the items in the measure to earlier sources that upon inspection do not appear to contain the actual or similar items. The complete CAQ measure is comprised of 45 items, 25 of which pertain to psychological abuse (the remaining items are included in order to balance the measure or because they pertain to physical abuse, also of interest to Gross and Keller). Examples of items included in the measure to asses psychological abuse include showing disinterest, belittling child, and name calling. Each item is rated on a 5-point frequency scale from never (0) to frequently (4), with the designated time period being childhood and adolescence. Items are rated for all parents/guardians combined. Total scores on the psychological abuse scale range from 0 (score of never, 0, on all 25 items) to 100 (score of frequently, 4, on all 25 items).

In the initial validation study, 260 undergraduates completed the measure, with an alpha of .85 for the 25 relevant items. A psychologically abused subgroup was identified through a complex coding system which resulted in about one fourth of the sample (54 of 228) meeting the criterion. Validity data were offered by way of the fact that the group of subjects who reported having experienced psychological abuse in their childhood reported statistically significantly higher scores on the Beck Depression Inventory. Although 50 other articles cite the Gross and Keller study, not one reports using the Child Abuse Questionnaire itself.

The Child Abuse and Trauma Scale (CATS) was created by Sanders and Becker-Lauson (1995) and contains 38 items, 7 of which were identified in a subsequent article by Kent and Waller (1998) as constituting a separate emotional abuse scale, including items such as did your parents ridicule you, did your parents insult you or call you names, and, as a child or teenager did you feel disliked by either of your parents. Respondents of the measure report on their experiences with both parents combined during their childhood and adolescence. Items are rated on a 5-point scale from never (0) to always (4), with total scores ranging from 0 (a score of never, 0, on all 7 items) to a score of 28 (a score of always, 4, on all 7 items). In Kent and Waller's (1998) validation study, 236 female undergraduates completed the measure, resulting in an alpha of .88. The mean score for the emotional abuse scale was .83 (SD = .86). This scale was significantly inter-correlated with the other scales on the same measure (r = .38 to r = .79) as well as with a measure of depression (r = .35) and anxiety (r = .38). Although over 30 articles cite Kent and Waller (1998), only a few actually employ the emotional abuse scale of the CAT. For example, Kennedy, Ip, Samra, and Gorzolka (2007) reported that emotional abuse had a direct effect on the disordered eating in a sample of college students. In that study, the mean score for the 7-item psychological abuse scale was 1.14 (SD = .69). The CAT itself is a widely used measure, although utilization of the emotional abuse subscale is much less common and there is no cut-off to differentiate maltreated from non-maltreated samples.

The Childhood Experiences Questionnaire (CEQ) was developed by Ferguson and Dacey (1997) to assess psychological abuse apart from physical or sexual abuse. The 30-item measure contained 11 items regarding psychological abuse, drawing from all five categories of the APSAC (Binggeli, Hart, & Brassard, 1981) definition. Each item was rated on a 4-point scale from never (0) to often (3). Total scores range from 0 to 33. The measure was validated in a community health center with 110 health care professionals, half of whom reported psychological abuse in their childhood (a score of 3 on any of the 11 items and no other form of abuse) and half who did not (no score of 3 on any of the 11 psychological abuse item and no other form of abuse as well). The two groups were compared on anxiety, depression, and dissociative experiences. Results revealed that psychologically abused women reported higher levels on all three outcomes measures. Although 26 articles cite Ferguson and Dacey (1997), none report using the measure itself. Thus, the only published article employing this measure is the original source article written by the measure's creators.

The Childhood Trauma Questionnaire, CTQ, is a 28-item measure (the short form) with 5 items related to emotional abuse and 5 items related to emotional neglect (Bernstein et al., 2003). Items ask about the subject's experience during childhood and adolescence and are rated on a 5-point scale from never true (1) to always true (5). Items are asked for all parents/adults combined and include, for example, being called names by family and parents wishing that the child had never been born. Total scores for each scale range from 5 (never, 1, on all 5 items) to 25 (always, 5, on all 5 items). In three different initial validation studies, the mean for each of the ten items was found to be between 1.8 and 3.0, although the scale means were not reported. Factor analysis, alphas, and inter-correlations have been presented to support the utility of the 28-item version of the CTQ, which is a widely used measure. For example the alpha for both the emotional abuse and emotional neglect scales were above .83 in a sample of adolescents, community adults, and drug abusers. Factor loadings were .71 or above for all of the items on both scales.

Subsequent publications reporting on the measure have presented cut-offs in which scores of 13 or greater are considered to reflect moderate to severe emotional abuse and scores of 15 or greater are considered to reflect moderate to severe emotional neglect (Rikhye et al., 2008). The CTQ is a widely used measure with over 1000 articles citing its use.

The Comprehensive Childhood Maltreatment Scale, CCMS, (Higgins & McCabe, 2001) is a 22-item measure with 3 items devoted to the experience of psychological maltreatment: yelled, ridiculed, or was cruel. Each item is answered on a 5-point scale from never (0) to very frequently (4). Respondents answer about their experience prior to the age of 13 separately for mother, father, and other. Total scores range from 0 (a score of never, 0, on all 3 items for all 3 possible perpetrators) to a score of 36 (a score of very frequently, 4, on all 3 items for all 3 perpetrators. In the initial validation study, the measure was administered to 175 community residents, which resulted in an alpha for the psychological maltreatment scale of .78. Mean scores for the scale were 16.68 (SD = 7.46) which was correlated with the CAT total score .84.

The measure has been used in 9 other published reports, mostly authored by Higgins and colleagues and the seminal article is referenced in 13 other papers. There is no cut-off to differentiate maltreated from non-maltreated samples.

The Conflict Tactics Scale (CTS-Parent–child Version) is a 22-item measure created by Strauss (1999). The premise of the measure is that all parents and children experience conflict of some form or another and that psychological aggression can be assessed by asking about whether parents respond to conflicts with nonviolent discipline (4 items), psychological aggression (5 items), minor physical assault (5 items), or severe physical assault (8 items). Examples of the psychological aggression items include: shouted/yelled/screamed, cursed/swore at, and threatened to send away or kick out of the house. Respondents rate the frequency with which each behavior occurred by each parent during a designated time period—usually when 13 years of age or the last year the individual lived at home. Items are rated on a 7-point scale from this never happened (0) to more than 20 times during the designated year (6). Scale scores for the psychological aggression scale range from 0 (score of 0 on all 5 items for each parent) to a score of 60 (a score of 6 on all five items for both parents). There is no definitive validation study utilizing the measure for adult recall and oftentimes the measure is used in a modified form. For example, Felitti et al. (1998) defined subjects in their study of adverse childhood events as having experienced psychological aggression if they responded “often” or “very often” to two items drawn from the CTS: (1) swear at, insult or put you down and (2) act in a way that made you afraid that you would be physically hurt. In that large scale study of over 9000 individuals, Felitti et al. classified 11% of the sample as having experienced psychological maltreatment. The CTS is a widely used measure in the field of marital conflict. It has been revised to also function as a measure of current parent–child relationships as well as adult recall of childhood abuse. There is no definitive cut-off when using the complete measure or just the psychological aggression items.

The Early Trauma Inventory—Self Report (ETI-SR) was developed by Bremner, Bolus, and Mayer (2007) as a self-report version of their clinical interview, the Early Trauma Inventory (Bremner, Vermetten, & Mazure, 2000). The ETI-SR is a 62-item questionnaire with items in four domains: general trauma (31 items), physical abuse (9 items) emotional abuse (7 items), and sexual abuse (15 items). The emotional abuse items include behaviors such as put down, ignored, and told was no good. Items are rated on a frequency scale but scoring is based on the number of items within each scale that are endorsed. Thus, the emotional abuse scale ranges from 0 (none of the items endorsed) to 7 (all 7 of the items are endorsed).

In a study of 278 adults, reliability was reported to be over .77. Endorsement of the 7 items ranged from 24% to 56%. Subjects were classified into one of five groups based on being a healthy adult (no psychiatric diagnosis) or having a psychiatric diagnosis. Those with borderline personality disorder had means higher than those with PTSD who had higher means than those in the abuse control and depression groups, which were higher than those who were classified as healthy adults. As a relatively new measure it has not yet been utilized by other researchers in the field, beyond the initial validation study conducted by the measure's authors.

The Exposure to Abusive and Supportive Environments—Parenting Inventory (EASE-PI) was developed by Nicholas and Bieber (1997). It is a 70-item measure with 19 items devoted to the issue of emotional abuse. Each item is rated on a 5-point scale from never (0) to very often (4) and rated separately for mother and father. Total scores range from 0 (score of 0 on all 19 items for both parents) to a score of 152 (score of 4 on all 19 items for both parents). Examples of items designed to measure emotional abuse include insulted/swore at, made to feel stupid, and treated like the black sheep of the family.

In the validation study, 255 undergraduates completed the measure. Factor analysis supported six separate factors: emotional abusiveness scale (19 items), physical abusiveness scale (13 items), sexual abusiveness scale (10 items), love/support scale (16 items), promote independence (6 items), and promote fairness (6 items). The emotional abusiveness scale was found to have an alpha of .84. Mean scores on this scale, which ranged from 0 to 76 were 10.2 for men and 15.3 for women and 13.0 for the sample as a whole. This is not a widely used measure with only 4 published studies reporting its use and no cut-off scores are available to differentiate maltreated from non-maltreated samples.

The Family Experiences Questionnaire (FEQ) was developed by Briere and Runtz (1988) to measure both physical and psychological abuse. The total measure contains 12 items with 7 items focused on psychological abuse, including yell at, insult, and criticize. Each item is rated on a 7-point scale from never (0) to more than 20 times a year (6). Items are rated for mother and father separately referring to when the respondent was 14 years of age or younger. Total scores range from 0 (a score of 0 on all 7 items) to a score of 42 (a score of 6 on all 7 items) for each parent. The validation study entailed administration of the measure to 278 undergraduate females, which resulted in an alpha of .87 for the 7 items for each parent. The two scales were correlated with each other .47.

This measure has been used in at least 10 other studies. For example, Reddy, Pickett, and Orcutt (2006) administered the FEQ in a study designed to assess whether a response style characterized by avoidance of negative private events mediated the relationship between reports of childhood psychological abuse and current mental health problems in close to 1000 college undergraduates. The FEQ has also been used under different names including Childhood Maltreatment Interview Scale (Clemmons, Dilillo, Martinez, DeGue, & Jeffcott, 2003), the Psychological Abuse Scale (Krause, Mendelson, & Lynch, 2003), the Psychological Maltreatment Scale (Varia & Abidin, 1999), and Psychological and Physical Maltreatment Scales (Weber & Cummings, 2003). These measures all used the same items and the same response options but vary the time period of interest (i.e., before age 17, prior to age 16, and an average year.) Although Briere and Runtz (1988) do not present a cut-off score, Clemmons et al. (2003) utilize a cut-off of 3.5 to differentiate maltreated from non-maltreated subjects.

The Life Experience Questionnaire (Gibb et al., 2001) is a 92-item measure with an 11-item emotional neglect scale and a 27-item emotional abuse scale. Examples of items include saying child is not as good as other children, punish unfairly, and humiliate or demean. Each item is rated on a 7-point scale from never (0) to over 20 times (7). Scale scores are created by summing the number of items within each scale. Thus, for the 11-item emotional abuse scale, scores range from 0 (no items endorsed) to 11 (all 11 items endorsed). All items are rated for the time period prior to age 15 and are coded for “caretakers” as opposed to any specific parent or guardian.

In the initial validation study, the measure was administered to 1204 participants in a larger study who were identified as either low risk for depression (n = 585) or high risk for depression (n = 619) based on cognitive styles and dysfunctional attitudes. Internal consistency for the psychological maltreatment scale was reported to be .85. Psychological maltreatment was found to be significantly related to depression in the sample as a whole and those in the high risk group had higher mean scores on the psychological abuse scale than those in the low risk group.

This measure is cited in 9 other published articles. No cut-offs are available for differentiating maltreated from non-maltreated samples.

The Negative Life Events Questionnaire (NLEQ) was created by Pitzner and Drummond (1997) and is a 74-item measure with 28 items devoted to what they refer to as psychological/verbal abuse (18 items) and control (10 items). Each item is rated on a 5-point scale from never (1) to very often (5). Items are answered in relation to a significant person committing the various acts during the respondent's childhood or adulthood. Thus, scores on the abuse scale can range from 10 (a score of 1 – never – on all 10 items) to a score of 50 (a score of 5 – very often – on all 10 items) and scores for the control scale can range from 18 to 90.

In a 3-stage validation study, various iterations of the questionnaire were completed by 95 residents of Perth Australia. Factor analysis was used to identify and confirm the three scales (two of which were related to psychological maltreatment) described above. Internal consistency for the abuse scale was .95 with a mean of 35.42 (SD = 13.26). Internal consistency for the control scale was .87 with a mean of 13.86 (SD = 5.13).The abuse scale was correlated with all of the criterion measures including paranoid ideation (r = .50), depression (r = .47), somatization (r = .28), and phobic anxiety (r = .28). The control scale was also correlated with phobic anxiety (r = .51). This measure has only been used in one other published study, also authored by Pitzner and Drummond (1997) and no cut-offs are available for differentiating maltreated from non-maltreated samples.

The Psychological Maltreatment Inventory (PMI) is a 25-item scale created by Engels and Moisan (1994). Each items is rated on a 6-point scale from this didn't happen (0) to extreme negative effect on me (5) and are rated as they relate to “parents” combined. A factor analysis of an original pool of 47 items resulted in 25 items falling into one of three factors: emotional neglect (12 items), hostile rejection (7 items), and isolation (6 items). Means and cut-off scores are not presented in the validation study, nor in any of the four articles in which the measure was also referenced. For example, Reyome and Ward (2007) administered the PMI to 122 nursing students and found that total scores for mothers and for fathers were each significantly correlated with scores on a co-dependency measure.

The Parental Acceptance Rejection Questionnaire (PARQ) is a widely used cross-cultural measure of the two key constructs of parental acceptance and parental rejection (Rohner & Khaleque, 2005). The measure has 60 items rated on a 4-point scale from almost always true (4) to almost never true (1). Each item is answered for mother and father separately. Examples of items include said nice things about me, paid no attention to me, and treated me harshly. Four scales are created from these items, warmth/affection (20 items), aggression/hostility (15 items), neglect/indifference (15 items), and rejection undifferentiated (10 items). Scale scores range from 20–80 for the warmth scale, from 15 to 60 for the aggression/hostility and neglect/indifference scales, and from 10–40 for the rejection scale. Total PARQ scores can range from 60 to 240.

Rohner and Khaleque (2005) report on extensive cross-cultural research supporting the reliability and validity of the PARQ. A meta-analysis of 43 studies of over 7500 respondents worldwide revealed statistically significant effect sizes for associations between parental acceptance/rejection as measured by the PARQ and subsequent adjustment/maladjustment (Khaleque & Rohner, 2002).

The Record of Maltreatment Experiences (ROME) was originally created and primarily used as a method for assessing maltreatment of children but has also been modified for use as an adult recall measure (Wolfe & McGee, 1994). In the adult recall version, the Family Background Questionnaire (FBQ), all 81 items are used, 17 of which refer to psychological neglect (e.g., offered comfort and reassurance to you when you were upset) and 24 relate to psychological abuse (e.g., ridiculed you or made fun of you in front of others). The items were based on the conceptualization of psychological maltreatment as, “communicative acts that have the potential to damage psychological attributes of the child” (e.g., self-esteem), whereas psychological neglect was defined as, “acts of omission that may result in negative psychological consequences for the child” (p. 169). The ROME is widely used, although the adult recall version, the FBQ, less so.

The Trauma Events Scale (TES) was created by Briere and Rickards (2007) to assess multiple forms of childhood abuse experiences, including a 20-item emotional abuse maternal figure scale and a 20-item emotional abuse paternal figure scale.² In a validation study, 417 participants completed the TES and an Inventory of Altered Self Capacities (IASC). Results revealed that emotional abuse by the mother – but not the father – was statistically significantly associated with (although the correlations were small) all but one of the seven scales of the IASC. To date, no other study has been published using this measure.

In addition to the development of these 15 scales, several researchers have measured psychological maltreatment with one or two items in larger batteries including Braver, Bumberry, Green, and Rawson (1992) who asked clients to check off from a list of types of abuse, including emotional abuse, which if any had occurred and which adult had been the perpetrator. No specific definition was provided to the respondents. Friedrich, Talley, Panser, Fett, and Zinsmeister (1997) included a single item in a large scale survey, “were you yelled at, threatened, or called demeaning names?” Respondents answered never, seldom, occasionally, or often for before and after age 13. Harter and Taylor (2000) asked respondents to complete a survey in which there were two items pertaining to emotional abuse, one a global endorsement of verbal abuse and the other pertaining to other types of emotional abuse. Hulme and Agrawal (2004) asked respondents whether anyone had “hurt your feelings or put you down a lot.” Kaysen, Scher, Mastnak, and Resick (2005) asked respondents three questions pertaining to emotional abuse, being called bad dumb or stupid, being threatened with a beating, or being cursed at. Responses were collapsed into a single dichotomous variable. Kernsmith (2006) asked respondents a single item about emotional abuse history (wording not available). Neziroglu, Khemlani-Patel, and Yaryura-Tobias (2006) used the New York State legal definition in their questionnaire, “non physical maltreatment of a child that can seriously interfere with his/her positive emotional development. The following behaviors constituted emotional abuse: constant rejection, terrorizing, exposing a child to corruption, violence, or criminal behavior, irrational behavior or verbal abuse (excessive yelling, belittling, and teasing.)” Rice et al. (2001) used a single item in their questionnaire in which respondents were asked if they were ever emotionally abused, defined as “made you feel bad through harsh words?” Sansone, in a series of studies used a single item, asking whether the respondents had been exposed to verbal or non verbal behaviors by another individual that were “purposefully intended to hurt or control you, not tease or kid you, before the age of 12?” (Sansone et al., 2001, Sansone et al., 2005, Sansone et al., 1998.) A single item was also used by Schneider, Baumrind, and Kimerling (2007), “Before the age of 18 did a parent or other adult in your household often or very often swear at you, insult or put you down or make you afraid that you would be physically hurt.” Titus, Dennis, White, Scott and Funk (2003) also used a single item, “abused you emotionally, that is, did or said things to make you feel very bad about yourself or your life.” Thus, not all researchers measuring this construct use multi-item scales designed to capture several different elements of psychological maltreatment.