Development and Validation of Parent Abuse Scale (Girl-Mother Version)

1. Background

Despite all social awareness and preventive and therapeutic programs, family violence in types of child abuse, partner abuse, and elder abuse is common in contemporary society. The other type of family violence is children violence toward parents named child-to-parent violence (CPV) that is attractive and its statistics are increasing. Cottrell (2001) named this family violence as parent abuse and defined it as “every action from child to create physical, emotional, or financial damage toward the parent in order to achieve power and control of them” (1, 2). Paglew (1984) used “parent abuse” for the first time in a discussion about children misuse toward their elder parent (3). In previous studies, the discrepancy in the parent abuse definition, research samples and measurement tools of this phenomenon has made more complex in the parent abuse literature (4).

This behavior is not just an important social problem, but it is the cause of researchers’ anxiety in different countries (5). The researches done in the USA showed that the violence toward parent is 7% to 18% in families with both parents, while it is 29% in families with a single parent (6). In a study by Margolin and Baucom (2014), it has been shown that 22% of sample adolescents had physical violence and 75% of them had verbal violence toward their parents (7). Calvete et.al. (2013) in a study on 1,698 Spanish adolescents indicated that 13.7% of this sample have at least one physical violence toward their parents (8). Almost all of them had one sort of psychological aggression toward their parents. The amounts of this behavior toward mother and father were 92% and 86%, respectively. This study illustrated that girls show more verbal and psychological violence toward their parent than boys. However, physical violence toward parents was more in boys’ behavior than girls. Mothers also were more abused than fathers (2).

Generally, there are some individual and social variables in the case of parent abuse etiology, such as family variables, individual features, previous violence in the family, cultural values, men power normality in cultures, etc. On the other hand, some social learning theories believe that parent abuse is a functional answer to family serious problems (9). According to this point of view, the effect of being exposed to family violence is the direct and indirect reason for parent abuse. One type of this violence is direct child abuse by parents. The indirect violence is spouse abuse and children are observer. Therefore, according to the intergenerational transmission of violence approach (10), it is assumed via observational learning and adult imitation model (11). Children in families with harsh behavior are possibly aggressive, thus they use aggression in confronting with individuals conflict. Children who observe violence between their parents may experience more parent abuse.

Studies have considered several main assumptions about parent abuse reasons: (1) Teenage prefer to face to conflicts with their individual way, but they simply repeat behaviors, which have learned because they cannot use their way; (2) When harmful behaviors occur, extreme excitements may overcome all family members; (3) Maybe both damaged person and damaging person feel disability and display their non-solved internal and external conflicts; (4) Teenage should recognize their annoyance experiences; and (5) Teenage can be helped in a change process by being equipped with alternative behavioral practices (3). Price (1996) has defined an interactive process that leads to parent abuse: (1) The adolescent has a request; (2) The parent ask him clearer information; (3) Adolescent answers politely and new answer will be created; (4) Although the parent confirms teen’s point of view, according to existing information, she prefers to say “no” to his teen, and she may cause a discussion with “another time” message; (5) The adolescent tries to change his parent’s decision by asking new question about his reasons. Sometimes in order to challenge his parent’s decision, he gives as much information to make sure his decision will not change (12). If the parent insists on his/her decision, teenage starts to say threatening word, the parent will get caught up, and finally, the teen answers by physical force, verbal threat, emotional misuse, and destruction of home appliances.

Prior studies have shown that there is a relationship between physical and verbal violence of parents toward children and physical and verbal violence of children toward parents (2, 13). The retrospective research has suggested that parents' violence toward children, especially mother's physical violence toward children, is the strongest reason for parent abuse (7). Adolescent abuse in the family has a direct relationship with his strong physical and psychological violence toward his mother (5). Beginning of harmful behaviors toward parents is more committed by boys than girls. The harmful behaviors toward parents by 10 and 11 years old boys and girls are similar. However, it is increased in boys older than 11 years old. The strongest parent abuse occurs by 11 to 17 years old boys, while it occurs by 10 to 13 years old girls, and is decreased after that age (14).

High emphasis on parental status in Persian culture is the other requirement to study parent abuse. The parental respect is very important. It is recommended to respect to parents in this culture. Iranian parents consider themselves responsible for the behavior and actions of their children and think that the mistreatment of children is a sign of their defect and impropriety. The other research has studied 364 high school boys and their mothers in Ahvaz (Iran) to develop and validating the parent abuse scale (boy-mother). Results of exploratory factor analysis using main factor analysis and orthogonal rotation (varimax) showed that the parent abuse scale (boy-mother) is made of financial, physical, and emotional factors. Convergent validity of parent abuse scale (boy-mother) is calculated through correlating this scale with child attitude toward mother (15). It indicated that the mean correlation coefficient of subject scores in the parent abuse scale (boy-mother) was significant with respect to child attitude toward mother (except for physical violence toward parents). Internal consistency of parent abuse scale (boy-mother) and its factors were calculated based on Cronbach’s alpha coefficient and were confirmed with correlation coefficients of 0.75 and 0.98.

2. Objectives

According to the previous studies and in order to complete the above research, no tools have been built to measure abusive behaviors of adolescent girls towards parents. Hence, this study is done to develop and validating parent abuse scale in adolescent girls.

3. Materials and Methods

3.1. Participants and Procedure

The statistical population of this study was all high school girls’ mothers in Ahvaz city (Iran). Two samples were used in this study. In the first sample, 409 people were selected for exploratory factor analysis and in the second stage, a random sample of 188 mothers was selected as a multi-stage cluster sampling for confirmatory factor analysis. The inclusion criteria were volunteer consent and the ability of reading and writing. The parent abuse scale (girl-mother) was given to participants in the first sample, after being insured of the agreement of the participants. We also asked mothers to give child’s attitude toward parent scale to students of mentioned high school, thus they could carefully answer the questions in privacy. The mean mothers’ age was 37.87 years with the range of 29 to 56 years, the standard deviation of 5.64, and the mean number of children was 2.12 with a domain of 1 to 5 children and standard deviation of 0.81 in the first sample. Moreover, 63.6% of mothers were housekeeper, 36.4% were employee, 11.7% had elementary school degree, 48.2% had diploma, 18.8% had associate degree, 16.6% were undergraduate, and 4.6% were postgraduate in the first sample. The mean mothers’ age was 38.17 years with the range of 30 to 55 years, the standard deviation of 5.74, and the mean number of children was 2.20 with a domain of 1 to 5 children and standard deviation of 0.88 in the second sample. Also, 62.8% of the mothers were housekeeper, 37.2% were employee, 10.6% had elementary school degree, 50.5% had diploma, 15.4% had associate degree, 18.1% were undergraduate, and 5.3% were postgraduate in the second sample.

3.2. Measurement

Parent abuse scale (girl-mother): The parent abuse scale (girl-mother) has 14 items, which is made for analysis and measuring two aspects of parent abuse (physical and emotional) in Iranian adolescent girls (Appendix 1 in Supplementary File.). This scale measures parent abuse by 5 degrees Likert scale (1 for never and 5 for always). Further two scores of parent abuse (physical and emotional), a score is measured for every subject in the range of 14 (min) to 70 (max). The highest score indicated high parent abuse.

Parent abuse scale (girl-mother) construction: (1) Parent abuse scale (girl-mother) construction’s objects determination, (2) Studying theoretical foundations and researches about parent abuse: presented theories and patterns, (3) Studying several researches that investigated parent abuse via interview, (4) Providing question treasury, (5) Obtaining face validity and content validity scale: to determine and provide parent abuse scale (girl-mother) content validity. This tool has given to 5 family experts and some changes and modifications were made in it according to their opinion, (6) Preparing a scale implementation instruction for subjects and examiners, (7) Random distribution of each question's components on a scale, (8) Build a key and scale scoring, (9) Determining the scale factor structure, and (10) Implementing final scale version and obtaining its psychometric features.

Child’s Attitude toward Parents scale (Hudson, 1992): It is a 50-item self-report scale (25 items for evaluating the severity of the child's problems with the mother and 25 items for evaluating the severity of the child’s problems with the father) that measures the severity of the child-parent relationship’s problems from child’s point of view. Each answer item is degreed on a 7-point Likert scale (1 for never and 7 for always). The items are classified as negative and positive to decrease the answer orientation. The scoring is reversed after executing positive items, and eventually, the items are summed up for child-mother and child-father. The high score indicated a negative attitude toward parents and severe problem in their relationship. Score 30 is a clinical cutting score. The scores below this point presented no important clinical problem in the relationship with parents. The Cronbach’s alpha coefficient of this scale was between 0.93 and 0.97 (16). The reliability of attitude scale toward mother and father were 0.85 and 0.86, respectively by alpha Cronbach method (17). twenty-five questions about child attitude toward mother were used in this study, and participant mother's adolescent girls completed the scale.

4. Results

Three main indicators were used to determine the number of factors: Eigenvalue, the ratio of explained variance by each factor and eigenvalue graph obtained via scree test. Figure 1 shows the extractive factors of parent abuse questionnaires. It can be seen from Figure 1 that two factors have eigenvalue more than 1.

The extractive factors of parent abuse questionnaires

Figure 1.

The extractive factors of parent abuse questionnaires

The factor analysis of the main components was first done using orthogonal rotation (varimax) on the scale items. The Kaiser- Mayer-Olkin’s measure of sampling adequacy tests (KMO = 0.89) and Bartlett’s test of sphericity (χ² = 5056.19) indicated that data had the factor analysis executable functionality. The Scree test also illustrated 2 factors. The factor analysis using orthogonal rotation to identify factor structure also certified scree test's results, showing that eigenvalue of two factors are more than 1 and explain totally 57.42% of articles variance. Item 13 was removed out from analysis due to the loading factor below 0.40. Items 1, 2, 3, 4, 5, 6, 9, 10, 11, 14, and 15 totally 46% variance and items 7, 8, and 12 and totally 11.42% variance were named “emotional parent abuse” in the first factor and “physical parent abuse” in the second factor, respectively. The emotional parent abuse (my child mocks me) and physical parent abuse (my child hits me when he is angry) are samples of extracted articles. Table 1 shows exploratory factor loads of parent abuse scale articles.

The mean statistical indicators and standard deviation of parent abuse scale articles are listed in Table 2. As can be seen, among all parent abuse scale items, item 1 has the highest (3.35) and item 15 has the lowest (2.80) mean, respectively. The correlation coefficient domain of each scale item is also flexible with a total scale score from 0.48 to 0.85, and is significant (P < 0.01).

As it can be seen from Table 3, the Cronbach’s alpha coefficient is 0.91 for the total scale, for two factors obtained from exploratory factor analysis-emotional parent abuse 0.93, and 0.75 for physical abuse. The simultaneous reliability coefficient of child attitude toward mother is achieved significantly 0.52 (P < 0.01), 0.27 (P < 0.05), and 0.54 (P < 0.01) for first factor, second factor, and total scale, respectively.

Table 4 shows a number of parameters and indicators of good fitness of the two-factor model of parent abuse scale. Fitness index of chi-square in the two-factor model was significant (χ² = 130.67, P < 0.01). It should be noted that with a correction index from e1 to e10, fitness indicators such as GFI and CFI were greater than 0.90 and the root mean square error of approximation (RMSEA) was less than 0.05, which showed a low rate of error in the model. In other words, it can be argued that a two-factor model with a correction in society is graceful.

Figure 2 shows the standard coefficients and factor loadings for significant factors. This figure shows that all coefficients of loading factor are significant in two factors.

Confirmatory factor analysis model for the two-factor structure of parent abuse scale

Figure 2.

Confirmatory factor analysis model for the two-factor structure of parent abuse scale

5. Discussion

Psychologist, criminologists, lawyers, sociologists, and other experts are investigating etiology and knowledge of familial violence in order to use the best interventions for the victims of this phenomenon. The study’s results confirmed psychometric properties of parent abuse scale (girl-mother version), including construct validity, convergent validity, and internal consistency in high school girl students’ mothers. The results of the exploratory factor analysis indicate that this scale consists of two emotional and physical parent abuse factors. These two factors totally explain 57.42% variance of parent abuse scale (girl-mother version). Also, 46% and more than 11.42% of the variance of total parent abuse scale belong to emotional and physical parent abuse, respectively. The correlation coefficient domain of each scale article was also variable between 48% to 85% with a total scale score and is significant (P < 0.01).

The Cronbach’s alpha coefficient is 0.91 for the total scale, 0.93 for emotional parent abuse, and 0.75 for physical parent scale. The simultaneous validity coefficient is calculated by correlating to child attitude toward mother. The total scale correlation coefficient is derived from child attitude scale toward parents as 0.54 (P < 0.01), first factor as 0.52 (P < 0.01), and second factor as 0.27 (P < 0.05). These results indicated that there is a significant positive relationship between parent abuse scale (total score) further two emotional and physical factors with child attitude scale toward mother.

The two-factor model extracted from this scale was tested by confirmatory factor analysis and it was shown that the chi-square analysis of this model was significant. It should be noted that the index is sensitive to samples above 100 people and not considered a valid indicator for model fitness by experts. However, this index showed that the present model has a good fit. Also, the RMSEA of this model was 0.07, which indicates a suitable fit in Iranian society.

In the last 60 years, our knowledge has limited development of what parent abuse is and why the parent abuse happens (4). Definition and recognizing “parent abuse” is the main limitation of this study. Defining “parent abuse” is difficult because it is unclear what behaviors are “normal” and what are “abnormal”. An important question is whether parent abuse is different from family violence. Holt (2016) believes that despite their some similarities, they have main differences. The other question is that who will define parent abuse (18). Cottrell (2005) presented that the experts -not families- can recognize the parent abuse behaviors of teenagers (19). The mentioned behaviors can be just misbehaving; however, this definition is not always consistent with the interpretation of the family of what is happening. Parents always know their children’s behavior as a normal one and relate it to the age of puberty or anything else. As most of the children have this behavior, parents believe that it is normal. Therefore, Simmons et al. (2018) argue that instead of focusing on their children’s age or developmental period, it is necessary to pay attention to the parent-child relationship (4).

The denial of the adolescent misbehavior seriousness is obvious when the parents tolerate the brutal behavior of the child without any action. The parents often tolerate these behaviors to support their children, hiding this issue and not engaging in judicial matters. Although “secrecy” can be seen in many forms of family violence, damaged parents deny their children’s behavior to (1) prevent next children's misbehavior, (2) support their family against social institutions that they do not trust them, and (3) support their self-image (3). Sometimes damaged parents suffer from depression, anxiety, and guilty feelings thus make a decision to deny this problem. This behavior from parents sometimes makes the adolescent think that their behavior is acceptable to them. Hence, the parents’ inability feeling becomes usual (20). In parent abuse and family violence case, the parents tend to hide this misbehavior (2). This issue can be a bit bolder in our culture. The parents anyway consider themselves responsible for all their children’s behaviors, and most of the time think that it is indicative of their inadequacy. Therefore, there is little statistics about the amount of parent abuse.

Cottrell (2001) defined parent abuse as any act from children to injure parents emotionally, physically, or financially in order to have more control on parents (1). This definition can be faced with some challenges regarding cultural and ethnic backgrounds. Now there is a question that whether it is really parent abuse when children do not want to make parents injure or control them. Although this definition has publicity and generality, there is no certainty that all these adolescents seek more control over their parents.

This study is done on mothers who participated in family education sessions in schools. Although all mothers are invited, we must be caution in generalizing the results of this research to other examples. The questionnaire about children also gave to them by their mothers and gathered them by mothers next session. Mother’s access to the answers of the questionnaire could make the children cautious to their responses. It is recommended to use other methods for investigating this variable due to challenges related to defining parent abuse. It also recommended investigating this variable with others relating to the parent-child field.