1. Background
Intimate partner violence (IPV) is defined as any behavior that can cause harm within an intimate relationship (1). IPV is linked to serious mental (2) and physical health consequences (3). According to a previous study, the prevalence of IPV across several countries was between 15% and 71% (4). IPV is also prevalent in Iran. One study reported a national prevalence of IPV of 66% in Iran (5), while another provided a rate of 94.7% in Tehran (6). IPV can be classified into several categories, namely, physical violence, sexual violence, emotional (psychological) abuse, and controlling behaviors (7). Generally, controlling behaviors are associated with violence against one’s partner (8), though little is known about the nature of this association (9).
Coercive controlling behaviors are prevalent. For example, one study reported that approximately 40% of men and women reported some form of coercive control (10). Moreover, research supports that coercive control impacts almost all areas of the affected person (11, 12). To our knowledge, no research has specifically investigated the prevalence and impact of coercive control in Iran. However, based on the literature on partner abuse in Iran, acts of psychological control, which can be similar to coercive control in nature, are common. For example, in previous studies in Iran, 37% of people reported restricting the employment of their partner (13), 52.7% reported being fearful of their husband’s rage, 23.6% reported being restricted by their husbands from leaving the house (14), and 47% reported that they were forced to do something unwillingly (15).
It can be challenging to define coercive control. Nevertheless, Johnson’s definition can be applied to conceptualize coercive control in relationships (16). Based on his definition, there are two types of partner violence: (1) coercive controlling violence, by which one person uses violent or non-violent tactics to coercively control his or her partner, and (2) situational couple violence, which results from the escalation of arguments and conflicts in a relationship. Because of the importance of coercive control in partner violence, it is crucial to have a valid measure to assess this kind of control. Previous research provided three facets of coercive control: (1) the abuser’s intentionality or goal orientation vs. motivation, (2) negative perceptions of controlling behaviors by the victim, and (3) the abuser’s ability to gain control through credible threats (17).
The Conflict Tactics Scale (CTS) has been used to assess tactics in a relationship (18). However, the CTS cannot assess controlling behaviors (19). Another common measure of coercive control is the Controlling Behaviors Scale (CBS) (20), which examines controlling behaviors in five subscales. Despite its ease of administration, the CBS only assesses behaviors (not intentions) related to control. In another study, the authors pinpointed that previous measurements of coercive control were generally unable to assess the motivation behind controlling behaviors, perceptions of the act by the victim, and the degree to which the perpetrator makes credible threats (17). Other researchers mentioned that the focus on actions rather than other aspects of coercive control, such as intentions, is a limitation of previous studies (21).
Other researchers used nine items from the Canadian Violence Against Women Survey and Psychological Maltreatment of Women Survey to evaluate controlling behaviors. In this questionnaire, individuals with three or more control tactics were considered highly controlling (22). This questionnaire is brief and easy to administer; therefore, we decided to assess the psychometric properties of this measure in Iran. However, we faced some challenges, such as the limited number of items used to assess coercive control. Accordingly, other items (for example, items related to checking one’s partner’s cell phone and threatening the partner harshly in cases of disagreement) were added.
2. Objectives
The present study aimed to extend the scale proposed by Johnson et al. (22) and provide a measure that fully represents the features of coercive control in Iranian couples. The psychometric properties of this measure were examined by evaluating its validity and reliability. Its factor structure was also investigated.
3. Methods
3.1. Study Design
This cross-sectional study was conducted between May and August 2021. All participants responded to the questions via Google Forms. All but nine participants lived in Iran at the time of the study. The present study is part of a larger project investigating partner violence in Iranian couples. The questionnaire proposed by Johnson et al. (22) was selected to assess controlling behaviors. Based on the items included in this questionnaire, it was named the Experiences of Coercive Control (ECC) Scale. Furthermore, a survey design was applied to assess the psychometric properties of the ECC scale.
3.2. Participants
The statistical population was all Iranian people (in Iran or abroad) who have been in at least one significant relationship in their lifetime and who have access to Google Form between May and August 2021. The sample size was chosen based on the ease of administration using Google Form. The sample size for principal component analysis (PCA) was chosen based on the recommendation to include a minimum of 10 subjects per item for factor analysis (23). Thus, the authors sampled 311 participants using the convenience sampling method.
3.2.1. Inclusion Criteria
To be included in the study, all participants needed to be Iranian and either married, in a relationship without formal marriage, divorced, or separated. In addition, participants needed to have at least minimal education and be able to read the Persian language.
3.2.2. Exclusion Criteria
Participants with major psychiatric or non-psychiatric disorders were excluded from the study. In addition, participants with a history of substance abuse (other than smoking and hookah use) or alcohol addiction were excluded.
3.3. Measurement Tools
3.3.1. ECC Scale
The ECC Scale is a 17-item questionnaire rated on a 5-point Likert scale (with responses ranging from 0 - 4). It was developed based on the items proposed by Johnson et al. (22). The internal consistency values of the scale for ex-husbands, ex-wives, current husbands, and current wives were 0.91, 0.83, 0.75, and 0.70, respectively.
One of the main authors of the present study translated the original questionnaire from English into Persian. The items were then back-translated into English and checked by a person with a full command of the English language. Some items were added to the questionnaire based on the authors’ opinions to ensure the scale fully represents the nature of coercive control in Iran. For example, the item “Tries to provoke arguments” in the original version was broken down into two items (“My partner shouts at me angrily” and “My partner swears at me in front of others”). In addition, the wording of some items in the original version was changed for consistency with the Persian language. For example, the item “My partner is jealous or possessive” in the original version was changed to “My partner is very sensitive to my behavior toward the opposite sex.”
After the modified version was developed, it was presented to 10 experts (two clinical psychologists, one health psychologist, one psychiatrist, and six Ph.D. candidates in clinical psychology) to assess the content validity. Based on the experts’ opinions, 17 items remained in the questionnaire.
The 17-item version of the ECC Scale was then presented to 28 individuals without any expertise in mental health, and the items were revised accordingly. All of the items, including those added to the original questionnaire, were subsequently back-translated into English by an English language expert and sent to Johnson (the designer of the original scale) for recommendations. Finally, the 17-item version remained for further analysis.
3.3.2. Wife Abuse Questionnaire
The Wife Abuse Questionnaire (24) is a 44-item scale of partner violence in Iran that includes three subscales: psychological maltreatment, physical maltreatment, and sexual maltreatment. The Cronbach α values for the total scale ranged between 0.87 and 0.95, and its estimated test-retest reliability was 0.98. The items related to sexual maltreatment were excluded because they were linked to women’s experiences. All other items, except those on the sexual maltreatment subscale, could be used by both women and men.
In the present study, the Cronbach α coefficient of this scale was 0.94. The confirmatory factor analysis results supported a three-factor solution. The convergent validity results were satisfactory, and there was a positive correlation between the results of the Wife Abuse Questionnaire and the Beck Depression Inventory (a measure designed to assess depression).
3.4. Study Procedure
The present study was approved by Iran’s National Committee for Ethics in Biomedical Research (code: 1006/98.ICBS.SBU.IR). Participants were recruited using Google Forms; the form received by participants included the research objectives, informed consent, and demographic data, as well as the ECC and Wife Abuse Questionnaire. The participants were selected through convenience sampling. Furthermore, in case any problems arose, the researcher’s contact information was available for consultation in the information section of the form.
3.5. Statistical Analysis
To assess reliability, the Cronbach α and interclass correlation coefficients (ICCs) were measured. In addition, to examine the construct validity, the Spearman correlation coefficient was measured. Moreover, to examine the factor structure, a PCA with a varimax rotation was performed. All analyses were performed using SPSS version 21 (SPSS Inc, Chicago, Ill, USA).
4. Results
4.1. Demographic Characteristics
The participants’ demographic characteristics are presented in Table 1.
Table 1.The Demographic Characteristics of the Participants a
| Characteristics | Values |
|---|---|
| Age, mean ± SD | 35.00 ± 8.46 |
| Gender | |
| Female | 271 (85) |
| Male | 47 (14.73) |
| Education level | |
| With academic education | 287 (90.6) |
| Without academic education | 29 (9.09) |
| Marital status | |
| Married | 272 (85.3) |
| Single | 18 (5.64) |
| Divorced or separated | 12 (3.8) |
| Others (not mentioned) | 17 (5.3) |
a Values are expressed as No. (%) unless otherwise indicated.
4.2. Reliability
The numbers of participants in the Cronbach α and test-retest reliability measurements were 311 and 29, respectively. The participants were assessed twice, with a two-week interval between tests. The Cronbach α was measured (α = 0.90) to assess the internal consistency of the ECC Scale. Furthermore, the intraclass correlation coefficient (ICC) and confidence interval (CI) were measured based on an average absolute agreement and two-way fixed effects model to assess the test-retest reliability. The ICC was estimated to be 0.96 (95% CI, 0.92 - 0.98). The Cronbach α and test-retest reliability coefficients indicate that the ECC Scale has high reliability.
4.3. Construct Validity
The sample size for the evaluation of the construct validity was 74 participants. The convergent validity was calculated using the Wife Abuse Questionnaire to assess the construct validity of the ECC Scale. Before the correlations were examined, the assumption of normality for the total scores of questionnaires was tested. The results of the Kolmogorov-Smirnov test for all three indices showed a non-normal distribution (P < 0.001). Therefore, the Spearman rank correlation coefficient was calculated instead of the Pearson correlation coefficient. There was a strong positive correlation between the ECC Scale and the Wife Abuse Questionnaire (n = 74; rs = 0.81; P < 0.001).
4.4. Factor Structure
To assess the factor structure of the questionnaire, PCA with a varimax rotation was performed. The Kaiser-Meyer-Olkin (KMO = 0.910) and Bartlett tests of sphericity (χ2 = 2917.975; df = 136; P < 0.001) demonstrated that the data were appropriate for PCA. The PCA yielded a 2-component solution, accounting for 57.34% of the total variance. The PCA results are reported in Table 2.
The results showed that all items of the scale had sufficient factor loadings. The details are presented in Table 3. In addition, the results revealed that there was a positive correlation between subscale control via aggression and spying behaviors, r = 0.447, n = 319, P = 0.001.
Table 2.Factor Loadings of the Experiences of Coercive Control Items
| Items | Factors | ||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | |
| 5. My partner humiliates me in public. | 0.706 | ||||
| 6. My partner criticizes me in public all the time. | 0.723 | ||||
| 7. My partner makes me feel helpless and incompetent. | 0.648 | ||||
| 8. My partner shouts at me angrily. | 0.789 | ||||
| 9. My partner swears at me in front of others. | 0.716 | ||||
| 10. My partner does not inform me about his/her earnings. | 0.695 | ||||
| 11. My partner does not allow me to use his/her earnings. | 0.787 | ||||
| 12. My partner controls my earnings. | 0.528 | ||||
| 16. My partner ignores me for a long time if I disagree with him/her. | 0.666 | ||||
| 17. If I disagree with my partner, he/she will behave so harshly that I do not dare to oppose him/her anymore. | 0.763 | ||||
| 18. My partner has made home an insecure and terrifying place for me. | 0.786 | ||||
| 1. My partner tries to restrict my communication with my family members and others. | 0.843 | ||||
| 2. My partner is very sensitive to my behavior toward the opposite sex. | 0.782 | ||||
| 3. My partner feels as if s/he owns me. | 0.773 | ||||
| 4. My partner always insists on knowing where I am and who I am with. | 0.703 | ||||
| 15. My partner always checks my cell phone. | 0.666 | ||||
| 19. My partner monitors all my commutes. | 0.653 | ||||
Table 3.The Subscales of the Experiences of Coercive Control Scale and the Statistics
| Factors | Number of Items | Mean ± SD | Skewness | Kurtosis | Cronbach α |
|---|---|---|---|---|---|
| Control via aggression | 11 | 8.73 ± 8.11 | 1.23 | 1.24 | 0.911 |
| Spying behaviors | 7 | 5.65 ± 5.15 | 1.17 | 1.21 | 0.855 |
5. Discussion
The present study is the first research conducted in Iran to assess the psychometric properties of a controlling behavior scale. The results support the validity and reliability of the ECC Scale, and the factor structure analysis yielded a two-factor solution. The first subscale of the ECC Scale (namely, control via aggression) consists of verbal aggression, threats, humiliation, criticism, economic control, gaining control by ignoring the partner for a long time, and diminishing the partner’s self-esteem. Overall, the first subscale seems to be related to cases of psychological aggression in which perpetrators use various tactics (such as verbal aggression) and severe acts (such as degrading, humiliation, threatening behavior, and isolation from others) (25).
Previous studies offer conflicting views about psychological aggression as a control strategy. Not all previous studies consider all psychological aggression tactics as controlling behaviors (26, 27). Thus, future studies should focus more on the relationship between psychological aggression and controlling behaviors.
Controlling behaviors have been reported as a significant predictor of psychological and physical violence (28-30). Thus, future studies should consider the unique role of these behaviors in different aspects of partner violence. One of the main features of coercive control is the abuser’s ability to control their partner using threats (17, 27), as mentioned in the present study. Including threats in the subscale of control via aggression also supports Johnson’s definition (16). Nonetheless, the nature of threats in relationships should be investigated in future studies. Finally, diminishing the partner’s self-esteem was included in this subscale as a feature of coercive controlling violence (16). The present study is consistent with previous studies concerning its attention to topics such as economic control (31) and humiliation (17). The second subscale of the ECC Scale (i.e., spying behaviors) comprises the following items: insistence on knowing the partner’s whereabouts, being sensitive about the partner’s behavior and attitude toward the opposite sex, monitoring the partner’s activities, feeling possessive of the partner, checking the partner’s cell phone, restricting the partner’s access to friends and family, and controlling the partner’s earnings. Many participants stated that both people in a couple should know each other’s whereabouts. There are apparent differences between partners who are informed about each other’s whereabouts and those who constantly insist on knowing where their partners are.
Moreover, the subscale of being sensitive about the partner’s behavior and attitude toward the opposite sex was called “being jealous of the partner” in the original version of the scale (16). However, because the concept of jealousy in relationships is ambiguous, future studies should further investigate the nature of jealousy and its manifestations in Iranian culture. Finally, controlling the partner’s earnings is an important factor related to economic control, which has been included in questionnaires, such as the CBS (32).
In addition, the current findings regarding the spying behaviors subscale are consistent with previous studies regarding the behavior of checking one’s partner’s whereabouts (33, 34), jealousy (35), possessiveness (36), and restricting one’s partner’s activities (37) as coercive controlling behaviors.
5.1. Conclusions
The present study had some limitations, including the relatively high number of female participants, the exclusion of the sexual maltreatment subscale from the Wife Abuse Questionnaire, and issues related to the sample size (e.g., sampling was done in different cities, and convenience sampling was employed). Nevertheless, the present study is the first to present a questionnaire assessing the level of controlling behaviors among couples in Iran. The ECC Scale is relatively brief and can be used in emergencies to assess whether a couple requires individual or couple therapy.
