Development and Validation of Attitude Towards Honour Killing Scale

Received 25 Feb, 2019

Accepted 10 Jun, 2020

Published 30 Jun, 2020

The present study aimed at developing a valid and reliable scale for assessing attitudes towards honour killing in Pakistan. The scale was developed in three phases; item pool generation, exploratory factor analysis, and confirmatory factor analysis. The initial item pool was generated form in-depth interviews with professionals (i.e., lawyers, journalists, psychologists, religious scholars, police officials, and social activists) and perpetrators in jail. In order to validate the initial 19 item scale, 459 adults, within the age range of 18-60 years were recruited from the Federal capital city and other cities of Punjab by using convenient sampling technique. For validation of the factor structure, Exploratory Factor Analysis (EFA) was run using Maximum Likelihood (ML) extraction method and promax rotation method. The analysis yielded two factors (affirmation and deterrents of honour killing) that accounted for 32% variance. Confirmatory Factor Analysis (CFA) was carried out to validate the factor structure explored through EFA. An independent sample of 695 adults was recruited for confirmatory study. Results of CFA indicated a good model fit for the final scale comprising 17 items. The Cronbach alpha coefficients for the two factors were .79 and .61, respectively. The convergent and discriminant validity of the final scale was also determined using the Gender Role Attitude Scale (Anila & Ansari, 1992) and Extremism Scale (Gilani & Altaf), 2005.

In an attempt to understand what honour killing entails it is imperative to develop an understanding of term honour. Honour has typically been described as a character trait associated with integrity and high sense of morality (Vandello & Cohen, 2003). However, a second less used definition of honour links it to reputation and status quo (Pitt-Rivers, 1966) and describes it as a socially achievable state that is determined by an individual’s capacity to uphold their will on others and theircontroltogainregardfrom others (Cohen, Nisbett, Bowdle, & Schwarz, 1996). Honour is therefore not only limited to personal pride and worth, but also associated with the position of one’s family in the society (Sakallı Uğurlu & Akbaş 2013). While, this definition is less likely to be used in western literature, it is a common conceptualization for a significant number of people all through the world and given its focus on the attainment and defence of honour, it is commonly employed when discussing violence for the sake of honour (Bowman, 2007; Vandello & Cohen, 2003).

In many patriarchal cultures, honour is a gendered notion, with women being repositories of honour and men inherently being its regulators and safeguards (Gupte, 2013). Following from this, honour killing can generally be defined as a form of a gender-based violence against an individual (typically a women) who is considered to have brought disgrace to the family or community by committing a sociallyobjectionable act (Cooney, 2014).

In several South Asian cultures, men are taught that their honour lies in family’s reputation, modesty of women, sexual pureness, and virginity of a women (Sakallı Uğurlu & Akbaş 2013; Vandello & Cohen, 2003). Men are deemed to have a sense of control over women’s social and sexual behaviour, with an inherent responsibility towards restoring their pride and honour whenever it has been challenged by their woman (Sakallı Uğurlu & Akbaş 2013). Often such notions of retributions infer that to restore the lost honour men must resort to killing the instigators of shame (i.e., their women) (Welden, 2010). Researchers have identified a number of methods employed for killing women in the name of honour, these include shooting, stoning, stabbing, and beating (Cohan, 2010).

Over the years, researchers have made efforts to shed light on the socio-psychological variables associated with honour killing. Within this realm, gender and gender role attitudes have been extensively studied (Esqueda & Harrison, 2005; Hillier & Foddy, 1993), followed by traditional gender role attitudes (Eisner & Ghuneim, 2013) and hostile and intolerant behaviours. It is quite apparent that incidents of honour killings are executed by the family members who are intolerant and harshly hostile towards the premarital intimate relationships, extra-marital affairs elopements, and secret marriages of their daughters. They consider these actions as an immense social dishonour which can only be recompensed by murdering the couple or a woman (Deol, 2014). Previous researches established that traditional gender role attitudes show positive relationship with favourable attitudes towards honour killing (Eisner & Ghuneim, 2013; Ilic, 2016). Therefore, present study has also selected gender role attitudes and hostility and intolerance variable to identify the relationship of these variables with newly developed attitudes towards honour killing scale.

Further causes attributed to the prevalence of honour; killings, include false accusations to settle personal conflicts and disputes under the pretext of honour; loopholes in judicial systems and parallel judicial system that is, Jirga (Jafri, 2008). People also misinterpret such cultural practices, especially, in Muslim majority countries, as being religiously sanctioned (Dogan, 2011). Additionally, the role of media in disseminating news about honour killings in a sensationalized manner promotes the practice, enabling other people to take similar actions against their women (Huda & Kamal, 2017). Finally, cultural tolerance and defense can be considered as a key factor in encouraging and maintaining the practice of honour killing (Cohan, 2009).

The roots of honour killing can be traced back to the cultural practices of the pre-Arab era; these practices are ingrained in the cultural fabric of Pakistan. Despite being labelled as a global phenomenon (Zia Ullah, 2010), honour killing is usually prevalent in Muslim majority societies (Shier & Shor, 2016). A study conducted in Turkey, a Muslim dominant country, revealed that when women behave contrary to family norms either by indulging into premarital sexual relations, friendships with opposite gender or extramarital affairs, their behaviour is considered as a smear on their family’s honour (Sev'er & Yurdakul, 2001). Similarly, a study showed that more than half of the Jordanian adolescents believed that honour killing is justified because maintaining integrity and honour is of utmost concern (Eisner & Ghuneim, 2013).

Pakistan, a Muslim majority country, has seen steady increases in reports of honour killings over the past few decades which are prevalent in every province of Pakistan (Warraich, 2005). Honour killing against a women is done for her alleged involvement in premarital or extramarital sexual relationships for the sake of honour restoration is practiced in all parts of Pakistan (Shaikh, Shaikh, Kamal, & Masood, 2010).

An important consideration when investigating honour killings is that despite it being a global phenomenon, its expression and public perception varies according to the culture in which it is being studied (Dogan, 2014; Rosenblatt, 2010). In the light of scarce literature available on honour killings in Pakistan, the present study conceptualizes honour killing as a killing of mostly a female member on accusation of her being involved in an immoral and socially inappropriate conduct (Ali, 2001; Faqir, 2001).

Understanding attitudes of general public towards honour killing are key in decoding the social-cultural triggers and context in which violence in the name of honour occurs (Copp, Giordano, Manning, & Longmore, 2016; Gracia, Lila, & European Commission, 2015; Waltermaurer, 2012). Given the imperative role public attitudes play in the prevalence and expression of honour killings, it is imperative to develop culturally ingrained tools to objectively measure the phenomenon.

Researchers have used a number of tools to assess people’s attitudes towards honour killings. Caffaro, Ferraris, and Schmidt (2014) assessed public attitudes towards honour killings in Italy and Turkey by adapting a questionnaire used by Golge, Yavuz, Müderrisoglu, and Yavuz (2003). They presented their participants with three scenarios pertaining instances of honour killings such as (honour killing for alleged adultery, honour killing for indulging adultery, etc.) and then asked them a set of standardized questions. The scenarios and questions were written up in Italian and translated/adapted to Turkish.

Another instrument to measure attitudes towards honour in Turkey was developed by Işık and Sakallı Uğurlu (2009). The Attitudes towards Violence against Women for Protecting Honour Scale assesses people’s attitudes towards honour and honour related verbal and physical violence against women. The scale comprises 14 items and records responses on a 6-point likert-type scale. However, the scale is only available in Turkish language and is part of unpublished undergraduate thesis.

A similar measure used to assess attitudes towards honour killings in Turkey was developed by Gürsoy and Arslan-Özkan (2014). The Attitude Scale for the Conception of Honour Related to Women (ASCHRW) has 25 items that assess three domains (traditional, egalitarian, and pre-marital sex) of public conceptions of what honour means with regards to women. The 5-point likert-type scale is designed to be used with young adults (aged 18 -24 years) and has not been validated for use beyond the Turkish culture. In addition, the development and validation articles on these scales are not available in literature to the best of researcher’s knowledge.

Another scale that aims at assessing honour killings amongst Jordanian youth was developed by Eisener and Ghuneim (2013). The Honour Killing Attitudes Scale (HKAS) comprises 4 items that asks respondents to rate the degree to which they think it is acceptable for a man to kill his sister, daughter or wife in the name of preserving honour and whether they think the concept is acceptable as a whole. The items were presented to the participants as a part of a 13-item set that presented a number of scenarios involving the murder of people. Unfortunately, the given scales have been constructed for the Turkish population and tested on the same. Moreover, these scales deal with concept of honour and do not fully cover phenomenon of honour killing.

Given the growing interest in honour killings in Pakistan, Rahim, Jahangir, and Zeb (2018) have developed an indigenous scale aimed at assessing attitude towards honour killing. This scenario-based scale comprises ten scenarios based on the information available in print and electronic media. However, it is important to mention that although the scale developed by Rahim et al. is relevant to Pakistani context, but, it has been published in 2018 when the researcher of present study had already developed a likert-type scale to measure attitude towards honour killing.

All the item statements of the present scale were generated through interviews of perpetrators and professionals who have worked on honour killings in their professional capacity. This had not been done in previously developed measures to the best of researcher’s knowledge. The statements in the scale could measure attitudes from multiple dimensions such as legal, religious, and sociocultural. Previous scales were mostly developed for measuring attitudes towards honour construct. Furthermore, few previous measure also lack reliability and validation process. The present scale of attitude towards honour killing is comprehensive in nature and fulfilled the requirement of development and validation.

Therefore, the focus of present paper is the development of a likert-type indigenous, valid, and reliable measure to determine the attitudes of individuals belonging to different walks of life. Likert-type scales are always beneficial when researchers are measuring latent constructs that is, characteristics of people such as attitudes, feelings, opinions, etc. (Babakus & Mangold, 1992). Despite the phenomenon being deeply embedded in Pakistani culture and society it is understudied due to the lack of objective measures.

Rationale of the Study
Pakistan has one of the highest reported rates of honour killing cases (Greiff, 2010). It is one of the important issues plaguing the society and ironically. Although, it is a global phenomenon, but it is more prevalent in countries where Muslim majority exists. The concept of honour has been explored by various studies, however there are very few likert-type scales designed to measure people’s attitudes towards honour killing. There is hardly any study done in Pakistan to reveal the attitude of people belonging to different social, professional, and educational backgrounds. It is important to determine attitude of diverse population to explore what kind of feeling and beliefs shape their attitudes.

Empirical research on honour killings is a relatively new dimension in Pakistan, especially, in terms of developing an understanding of people’s attitudes towards honour related violence and killings (Khan, 2018). Literature has established that it is important to measure attitudes and beliefs because they are strongly linked with is behavior (Eagly & Chaiken, 1993). Moreover, people also take influence from other people attitudes and beliefs and try to act in accordance with them, to protect honour of their family.

Therefore, present research (Study 1 and Study II) has been conducted with an aim of developing a standardized validated measure for honour killing to explore and determine the attitude from a multidimensional lens multidimensional (i.e., law, media, society, religion, and family). It has been presumed that significant change cannot be brought without identifying the opinion of people. The current study aims to fill this gap in the literature by giving a reliable and valid measure.

Objectives of the Study
Present study was carried out to achieve the following objectives.

1.	To develop a standardized indigenous measure of attitude towards honour killing
2.	To establish a psychometric properties of the Scale through factor analyses and convergent and discriminant validity with modern gender role attitudes and hostility and hostility/ intolerance

METHOD

The Attitude Towards Honour Killing Scale (ATHKS) was developed and validated in two studies. Study 1 was aimed to develop scale and establish its factor structure. Study 2 was aimed at establishing psychometric properties.

STUDY I

Study 1 was completed in two phases. In Phase 1, items were generated based upon in-depth interviews; and in Phase 2, data were collected on initial form of the Scale and factor structure was established through exploratory factor analysis (EFA).

Phase 1: Item Pool Generation
Sample. Purposive sampling was employed to collect the data. In Study 1, the item pool was developed primarily based on categories and themes identified through in-depth interviews (N = 30) with professionals were conducted. Sample included lawyers (n = 4 male, n = 3 female) social activists (n = 6 Females, social activists also included NGO workers who are working on women related issues), journalists (n = 3 male, n = 2 female), religious scholars (n = 4 male) psychologist (n = 3 females) police officials (n = 3 males, n = 2 female) and perpetrators of honour killing (n = 5) in jail setting. Interviews were conducted till the saturation point. The inclusion criteria for sample selection was that only those professionals were contacted who had an experience in dealing with honour killing cases in their professional capacity, for example, only those journalists were contacted who had an experience of reporting honour killing cases; lawyers who had dealt with honour killing cases; religious scholars and psychologists who had their researches on honour killings; and only those perpetrators were interviewed in the jail who were convicted for killing in the name of honour. For perpetrators’ interviews, permission was taken from Inspector General prison, Punjab, Pakistan. Finally, these professionals were selected for the interviews based on their experience related to honour killings in their fields.

Interview Guide. In-depth interviewing technique was employed to collect the data. Interview protocol was formulated with the help of observations, previous literature, and through focus group discussions conducted with graduate students, general public including men and women, and a focus group with low socio-economic background (such as security guards, housekeepers, drivers). It is important to mention that people from low socio-economic backgrounds and with minimal education were interviewed, because it was observed and also reported in Pakistani newspapers and other forms of literature that honour killings are more prevalent amongst people from low socio-economic status. Therefore, Interview questions were finalized by incorporating questions based on literature, researcher’s observation, and themes identified during focus group discussions. Open-ended questions with probing questions were developed to ask questions such as: “What is your understanding about the phenomenon of honour killing?”, “How do you believe honour killing stems out of our own social structure?”, “Why does Honour matter a lot in our society?”, and “What is your opinion about the present law dealing with this crime?” etc.

Procedure. All the interviews were audio-recorded with the permission of the participants of the study. However, interviews with perpetrators were recorded in a written format as recording was not allowed by the jail authorities. The interviews length varied from 45 minutes to 75 minutes. All audio-taped interviews were later transcribed by the researcher. The grounded theory analysis was employed to identify and analyse codes from the transcribed interviews. Similar codes were clustered into categories and then further clustered to create analytical categories. The item pool for the scale development was made through codes identified in the interview phase.

Initial form of the ATHKS. Total 37 statements were generated based on the frequently occurring codes and categories from interviews. The core categories identified in the qualitative phase shed light on following aspects that is understanding and conceptualization of honour killing and institutional factors (family, religion, media, law, police, and judiciary) that contribute in formulating attitude towards honour killing. The developed item statements were reflective of both positive and negative beliefs about honour killing. Such as statements like: “Societal and family pressure is far more important than the life of a girl”. “Law should be relaxed towards killings in the name of honour, etc.” High scoring on these statements would reflect affirmative attitudes and acceptance of honour killing among individuals. Whereas, statements like: “As per state law killing in the name of honour should be punished like any other murder. “In case of honour killing a murder should not be given advantage of forgiveness despite of being a family member.” High score on these statements would be reflective of deterrent attitudes and disapproval towards honour killing by individuals.

All the initial items were critically reviewed by the first and second author to address comprehension, redundancy, complexity, ambiguity, and cultural relevance. Finalized statements were selected for the further review from Subject matter Experts (SMEs). Response format of five-point likert-type was also decided in this phase (Strongly Disagree to Strongly Agree). Likert-type scales may meet the research needs when researchers are interested in measuring attitude, belief, or behavior items. A five-point likert scale was used to increase response rate and response quality along with reducing respondent’s frustration level (Babakus & Mangold, 1992). Content validity of ATHKS was assessed through the careful selection of items from panel of experts who evaluated the relevance of the items to capture the key characteristics of the construct. The committee of five experts included three Assistant Professors and two PhD scholars of Psychology. These SMEs were approached based on their expertise in psychometrics and gender related issues. They were requested to review all the potential items carefully. Consequently, statements reflected difficulty in comprehension; vagueness, and overlapping with other statements were dropped. Out of 37 items, 19 items were selected for empirical evaluation as initial form of ATHKS through EFA.

Phase II: Exploratory Factor Analysis (EFA)
EFA was carried out to explore the factor structure of the Scale. Before conducting, EFA the inter-item correlation analysis was conducted that ranged from .10 to .47 at p<.05, .01.The weak to moderate inter item-correlation is an evidence of non multicollinearity of the data

Sample. Participants (N = 459) were men (n = 245) and women (n = 209) with age range of 18-60 years (M = 28.56, SD = 10.12). Moreover, 5 participants have not mentioned their gender. Data were collected from Federal region and different cities of Punjab, Pakistan namely, Lahore, Sargodha, Chakwal, Rawalpindi, and Multan. The inclusion criterion for recruiting the sample was minimum age of 18years and minimum education to be matric. People below age range of 18 years and under matric were not included in the study.

Procedure. Data were collected by using 5-point likert-type scale (1 = Strongly Disagree to 5 = Strongly Agree) for 19 items of initial form of ATHKS and demographic sheet through convenient sampling technique. Participants were ensured about the confidentiality and their voluntary participation was recorded via informed consent. They were also debriefed about the purpose of research; that is, to develop a measure to measure attitudes of people towards honour killing. Permission for data collection was also taken from institutional review board. Permission was also taken from the heads of respective universities for the students who took part in the study. Moreover, permission was also sought from Inspector General (IG) prison Punjab, Pakistan for the data collection of perpetrators. EFA was carried out on the collected data through SPSS 21.

Results. Before carrying out EFA, few tests were applied for the authentication of data fitness for the factor analysis that is Barlett Test of Spehericity (χ² = 1590.33 (231), p < .001), and Kaiser-Meyer-Olkim (KMO = .80) values were computed. Along with these tests, factors based on Eigen value >1were retained for the further analysis. Any item with factor loading less than .35 and items who cross-loaded on other factors were not consider for further inclusion in final form of ATHKS. Items were finally included on the basis of interpretability and meaningfulness of factors in the context of research.

EFA was conducted by using maximum likelihood (ML) extraction method and promax rotation method. Schmitt (2011) strongly endorsed that researchers must use method of oblique rotation, because it tend to give simpler statistical structure and realistic results. Item-total correlations (.10 to .47 at p < .01) indicated that items can be tested through promax rotation. Five factors emerged with eigen values more than 1. Out of five factors, three showed variance less than 5% except for two factors which explained 21% and 11%, variance.

Factor solution was tried by fixing number of factors (4, 3, & 2) and were analyzed based on their item loadings (> . 35), face validity of items in each factor, and overall content of the factor. Four and three factor solutions revealed lack of meaningful cluster. Repeated analyses suggested two factors solution as the most appropriate solution for ATHKS. In two-factor solution, the variance explained by factors was 21% and 11%.

Item loadings, eigen values, and variance percentage explained by the factors are mentioned in Table 1. Seventeen out of 19 items are retained after EFA with satisfactory item loadings (> .35). Results show communality of mostly items > .35 which is indication of less specific difference among variables. The two items were deleted because they have less item loading than the set criteria. Additionally, results direct that Factor 1 has an eigen value of 4.08 and explains 21.41% of the total variance whereas, Factor 2 has an eigen value of 2.09 and explains 11.01% of the total variance. The Table 1 demonstrates that total variance explained by two factors is 32.43%.

Two factor solution was retained on the basis of SMEs opinion variance explained by each factor, their interpretable meaningfulness, and relevance with the factor.

Table 1:	Factor Loadings of Items of Attitude Towards Honour Killings Scale (ATHKS) (N = 459) of the study areas

The items falling in the two factors were given to five research experts (PhD in Psychology). The experts were requested to label both subscales with appropriate names based on the face validity of items falling in each subscale. ATHKS was finalized with 17 items with response categories 1 to 5 (Strongly Disagree to Strongly Agree). Original scale is in Urdu language. However, scale has been translated in English language for foreign readers/researchers. The translation of scale has fulfilled all the standard psychometric requirements.

Factor One: Affirmation of Honour Killing (AHK). Overall 11 items (1, 2, 4, 5, 8, 10, 13, 14, 15, 16, & 17) loaded on this factor (see Table 1) indicating more affirmative and accepting of honour killing. Therefore, high scores indicated affirming or endorsing honour killing and low scores indicated less acceptance of honour killing. Sample items are “There should be no punishment for honour killings in the law.” “Killing is the only way to deal with anyone who acts against Islamic law (shariah).” The score range for AHK is 11-55.

Factor Two: Deterrents of Honour Killing (DHK). Overall 6 items (3, 6, 7, 9, 11, and 12) loaded on this factor which indicates deterring attitudes and non accepting attitude towards honour killing. Therefore, high scores indicated not favouring honour killing as a strategy to control morality or protect honour. The score range for DHK is 6-30. Example items are “Lack of punishment for honour killing in the state law encourages people towards it”, and “As per state law, killing in the name of honour should be punished like any other murder.”

Both the subscales are scored independently. Present study have not aimed to take composite scores, however, it could be taken by reversing the items of any one of the subscale.

STUDY II

Study II was carried out to establish psychometric properties of ATHKS. Confirmatory factor analysis (CFA) was done along establishing convergent and discriminant validity of the Scale.

Hypotheses were established to assess convergent validity for both the subscales of ATHKS with modern gender role attitudes and hostility/intolerance (Extremism Scale).

1.	Affirmation of Honour Killing would positively correlate with Hostility/Intolerance Subscale (Extremism Scale) and negatively correlate with modern gender role attitudes
2.	Deterrents of honour killing would positive correlate with modern gender role attitudes
3.	Deterrents of honour killing would show zero correlation with Hostility/Intolerance Subscale (Extremism Scale)

Sample
Participants (N = 695) of the study included men (n = 248) and women (n = 447) with age range of 18-60 years (M = 25.89, SD = 8.10) who responded on a 5-point likert-type scale (1 = Strongly Disagree to 5 = Strongly Agree) for 17 items ATHKS, along with other measures to study convergent validity of the Scale newly developed inst. Data were collected from Federal region and different parts of Punjab. Inclusion criteria were same as of Study I.

Instruments
Gender Role Attitude Scale. The 5-point likert scale comprises 32 items developed earlier as Sex Role Attitude Scale (Anila & Ansari, 1992), and was later revised and translated by Kamal and Saqib (2004). The revised and translated scale comprised of 30 statements which was used in the current study. The scale measures roles of men and women in different categories including inside and outside home, occupational activities, personal relationship, academic achievements, decision making, and for parental responsibilities. The two subscales of the instrument include modern and traditional attitudes. As the two subscales are opposite in nature, items of Traditional Attitudes (1, 3, 5, 8, 9, 14, 15, 16, 17, 21, 24, 25, 28, 29, & 30) were reversed to measure modern attitudes on the scale. The reliability of the scale as mentioned by authors was .87. Gender Role Attitude Scale (Anila & Ansari, 1992) was also used to establish convergent validity of the ATHKS.

Extremism Scale (ES). It was developed by Gilani and Altaf (2005). In order to measure the extremist attitudes in the sample, the Extremism Scale (Urdu version) was used. The scale consisted of 42 items with five subscales namely; Conservatism (16 items), Hostility and Intolerance (8 items), Submission to authority (7 items), Rigidity (8 items), and Power and Toughness (3 items). The 5-point likert scale had scores ranging from 1(Strongly Disagree) to 5 (Strongly Agree). Only Hostility/Intolerance (2, 5, 10, 12, 18, 20, 29, & 37) was used for the current study to establish convergent validity of ATHKS. High scores on Hostility/Intolerance Subscale indicated the emotionally charged angry behaviour of individuals to express the conduct of indignation and animosity.

Procedure
Same procedure was followed for Study II as of Study I.

RESULTS

Confirmatory Factor Analysis (CFA) of ATHKS. CFA was conducted by using Analysis of Moment Structure (Amos 22) statistical package. CFA with maximum likelihood was carried out to test whether the factor structure of ATHKS could be replicated on an independent sample. CFAs are carried out to test the hypothesis (Stevens, 1996), and the current paper proposed that the factor structure of ATHKS would remain the same on an independent sample as well.

In order to ensure the factor structure of the ATHKS and its subscales that is, Affirmation of Honour Killing and Deterrents of Honour Killing. The Model 1 of the present study (default model with all the 17 items) showed all the indices in acceptable range such as CMIN/df< 3; CFI = .91; GFI = .95; TLI = .89, RMSEA = .04 with χ² = 289.8 (118), p< .001. Therefore, no modification indices were applied; hence, CFA indices confirmed the current model.

Table 2 presents all factor loadings of the items within set criteria (< .30). Therefore confirms 17-item factor structure for ATHKS, Figure 1 represents the pictorial representation of the model.

Fig. 1: Measurement model of ATHKS; F1 = Deterrents of Honour Killings; F2 = Affirmation of Honour Killings

Figure 1 represents the graphical picture of a model. The factor loadings in the model ranged from .32 to .59 for Affirmation of Honour Killings Subscale for item no. 17 to 10, respectively. The factor loadings in the model ranged from .36 to .58 for Deterrents of Honour Killings Subscale for item no. 6 to 3 respectively.

Descriptive Analysis, Convergent and Discriminant Validity
Table 2 shows descriptive characteristics of the data. The values of the descriptive suggest that the data fit on the normality assumption. Reliability coefficients of the variables are acceptable except for Deterrents of Honour Killings and Hostility/Intolerance Subscale of Extremism Scale. Low internal consistency of Deterrents of Honour Killing scale and hostility and intolerance scale can be explained in terms of less number of items.

In this phase, evidence of convergent validity of Affirmation of Honour Killing and Deterrents of Honour Killing subscales have been established with Hostility/Intolerance Subscale of Extremism Scale (Gilani & Altaf, 2005) and Gender Role Attitudes Scale (Anila & Ansari, 1992), respectively.

As expected, moderately positive and significant association is obtained between Affirmation of Honour Killing of ATHKS and Hostility/Intolerance Subscale, whereas, the results show significant negative correlation between Affirmation of Honour Killing and modern gender role attitudes. Furthermore, results also indicated significant positive correlation between Deterrents of Honour Killing and modern gender role attitudes. Thus these results were evidence of convergent validity. Furthermore, Deterrents of Honour Killing and hostility/intolerance showed no correlation and indicated discriminant validity.

Table 2:	Correlations Between Attitude Towards Honour Killing and Its Subscales (Affirmation and Deterrents), Gender Role Attitudes, Hostility/Intolerance (N = 695)
Note. S= Skewness; K= Kurtosis **p < .01

Study II confirmed the factor structure of the indigenously developed ATHKS on an independent sample. Moreover, convergent validity was also established with gender role attitude and hostility/ intolerance (Extremism Scale) as hypothesized.

DISCUSSION

Conceptualization of honour was found to be a definable factor that drives towards honour killing. Women have to bear the responsibility of their own honour as well as of their family members. Thus women as compared to men, who breach the honour codes set by the family and for which society will have to face extreme consequences (Reddy, 2008). Comparable to the results of this study, researchers (such as Reddy, 2008; Welchman & Hossain, 2005) concluded that the preservation of honour is inseparably aligned with masculinity. Women are mostly seen as victims in honour killing incidences because the concepts of honour and shame are largely associated with female sexuality. Nonetheless, Pakistan has similar perception for honour as stated above which leads to the prevalence of honour killing. Therefore, it has been reported extensively in Pakistan in past decade and its roots can be traced back in history as well (Jamali & Shah, 2015).

Honour killing in Pakistani context has been understood as killing of mostly a woman for bringing shame and dishonour to the family by being involved in culturally deviant behaviour (Ali, 2001). Based on the previous findings people at large carried favourable or positive attitude towards honour killing (Shaikh et al., 2010; Shaikh, Kamal, & Naqvi, 2015; Rahim, Jahangir, & Holden, 2016). It has been reported in Pakistani literature that men carried more positive attitude than women towards honour killing (Rahim et al., 2016). Similar type of studies conducted in other countries like Jordan on Muslim sample also showed that adolescent boys showed more favourable attitude towards honour killing than adolescent girls.

The present study is aimed to construct a psychometrically sound scale that is capable of measuring attitude towards honour killings. The statements reflected the beliefs and feelings of people towards honour killing and were designed to assess an overall spectrum of attitude towards honour killing. Perhaps there is no indigenous likert-type scale available to measure people’s attitude on honour killings with multi-dimensional perspective. Few items from both subscales of ATHKS are mentioned above for better understanding that how it measures attitudes on multi-dimensional level. The original scale was developed in Urdu language so that it could be administered on Pakistani population at large. However, items were also translated for non Urdu speaking people through following guidelines of standard translation process by (Gudmundsson, 2009).

The instrument can be tested on people with minimum education of matriculation and with minimum age of 18 years. The validation process employed using both EFA and CFA. EFA is used when researchers have slight idea about the primary mechanism of the target issue; thus are unsure about how the variables will function with each other (Henson & Roberts, 2006). EFA is a medium to assist build a new theory by investigating latent factors that best account for the differences and interrelationships among established variables (Henson & Roberts, 2006). Total variance explained by two factors is 32.43%. Although no absolute threshold is defined for cumulative variance, nonetheless, it is proposed that higher cumulative variance yields better factor solution, yet, in social sciences cumulative variance below 50% variance is also acceptable (Williams, Onsman, & Brown, 2010). One possible explanation for the low cumulative variance is homogeneity of sample (Field, 2009).

Item generation for present study has been conducted with the help of categories and subcategories that emerged in the qualitative phase. To explain the process briefly, in the qualitative phase, semi-structured interviews were conducted with the perpetrators, professionals such as lawyers, police officials, journalists, social workers/activist, religious scholars, and psychologists having professional experience related to honour killings in their fields. After conducting EFA (N = 459), 17 items were selected for the CFA. Factor analysis was conducted for the development of the scales. Items in the scale were finalized using EFA (ML). The inter-item correlation among item statements provided an evidence to use oblique rotation. It is suggested to use this rotation method when items in the scale are correlated with each other. The reliability of the honour killing subscales showed that both the subscales have satisfactory reliability, therefore, the measure is reliable to use for further studies. The two factors that emerged in the exploratory phase were reflective of two types of attitudes; 11 items represented affirmative attitudes towards honour killing were loaded on factor one, whereas, six items that were reflective of deterring attitude were loaded on factor two. Therefore, items of both the factors were scored independently. Previously, in Pakistan no such scale has been developed that measured attitudes in accordance with two dimensions.

Confirmation and validation of factor structure of newly developed ATHKs was done through CFA. CFA was carried out on another dataset of 695 individuals. All the indices of CFA revealed an acceptable model fit to the data. All the items loaded independently on their respective factors. The results of reliability analysis showed that the scale is reliable and consistent to use further. Items-total correlations were also significantly and positively correlated with the total of ATHKS. However, low internal consistency of Deterrents of Honour Killing subscales can be because less number of items. The evidence in the literature argued that threshold for Cronbach’s alpha values found in the literature is usually equal or superior to .70 (Nunnally, 1994; Pasquali, 2003). Yet values around .60 are acceptable when the instrument is used for research purposes and not for diagnostic purpose (Hair, Anderson, Tatham, & Black, 1998). Previous scale constructed to measure the concept of honour in collectivistic societies have also reported low internal consistency in their subscales such as Cronbach’s alpha of each dimension of Honour Scale (revised version developed by Guerra, Gouveia, de CR Araújo, de Andrade, & Gaudêncio, 2013) was .77 (family honour), .66 (social honour), .82 (feminine honour), and .61 (masculine honour).

Additionally, the study has established the convergent and discriminant validity of newly developed ATHKS with the subscale of Extremism (i.e., Hostility/Intolerance) and modern Gender Role Attitude respectively. It has been discussed earlier in the literature section as well that the gender role attitude is crucial in understanding honour killing. Gender role attitudes are broadly of two types, conservative/traditional and modern/egalitarian. Henceforth, it has been widely discussed in the literature that traditional and conservative gender role beliefs favour killing in the name of honour (Eisner & Ghuneim, 2013). Moreover, the study conducted by Ilic (2016) also confirmed that traditional mind-set and attitudes endorsing the legitimacy of honour killings. Therefore, in an attempt to understand attitudes towards honour killing, it was important to closely observe this phenomenon in regards to gender role attitudes.

In the light of above mentioned factual findings of this study duly supported by literature on the subject, it can be established that, when a woman contravenes her gender roles by acting against the gender expectations of the society and family, it causes serious threat to male family members. Consequently, the restoration of honour is possible through killing (Wikan, 2008). The man’s respect is, hence, established in his capacity to secure and control his family, especially, with respect to gender-normative roles (Wikan, 2008). It is evident through multiple findings in the literature that prescribed gender role attitudes serve as strong predictor for honour killing attitudes as well. Therefore, by estimating the importance of gender role attitude variables with regards to honour killing researches, it has been selected for establishing convergent validity.

Thus, as hypothesized, the findings indicated that modern gender role attitudes turned out to be negatively related with affirmative attitude towards honour killing. The negative correlation among them is because people who carry modern gender role attitudes will be less tolerant towards honour killing, whereas, results also indicated positive correlation among deterrents of honour killing and modern gender role attitudes. This would signify that, the more people have egalitarian attitudes towards women the lesser they would assert the idea of honour killing. Literature also guides with similar findings of study conducted on both male and female residents of Barbados society in which the study identified that the residents who had traditional gender role attitudes perceived the violent action against the female victim as justified (Linton & Lorde, 2011). Similar findings have been identified in the past researches, which claim that people having more traditional attitudes will be more prone towards hostile sexism (Hill & Marshall, 2018). Consistent with the current findings another study also reported that men believing in strict patriarchal traditionalism showed strong favour towards honour killing phenomenon (Eisner & Ghuneim, 2013).

Honour killing is considered as a violent practice. It has been considered an extreme form of violence which cannot be carried out without having extremist attitudes (Grzyb, 2016). This ensures that hostility/intolerant attitudes among individuals motivate them towards violent actions. Literature confirmed that people with personality traits such as hostility and impulsivity are more at risk of engaging in criminal behaviour. Nonetheless, honour killing is an act of extremism. Thus, the variable of hostility/intolerance has been selected to observe its relationship with honour killing attitudes. Literature from various parts of the world has confirmed that hostile traits and cognitions do play role in honour killing attitudes. As one of the study showed that in the past two decades, honour killings have accelerated considerably.This may indicate that honour killings are genuinely growing, perhaps because of fundamentalist and extremist attitude (Chesler, 2010).

Thus, present study also showed consistent results with previous literature. The outcomes indicated positive correlation between extremism (hostility and intolerance) and affirmative attitudes towards honour killing. It can be presumed from the findings that those having hostile/intolerant attitudes are a great risk factors for indulgence in honour killings. It has been established with the present findings that the hostility/intolerance correlate with affirmation of honour killings. People who have hostile and intolerant emotions and attitudes generally towards different approaches of life are more likely to choose for extreme actions or vice versa. Similar findings had been observed where hostility and intolerance has been reported as positively associated with intimate partner violence (Elmquist et al., 2016).

This study showed no correlation between deterrents of honour killing and hostility/ intolerance variables, giving the evidence of discriminant validity. Hostile and intolerant people are least likely to have a negative attitude towards honour killing. Their idea of extreme reactions contradicts with the benevolent nature of showing deterrence towards honour killing. However, the results of present study had shown neither positive nor negative correlation. However, studies on similar construct showed negative correlation between positive/ affirmative attitude towards honour killing and extreme attitudes such as obedience and aggressive behaviour (Eisner & Ghuneim, 2013; Ilic, 2016). Literature has indicated that people whose attitudes are not in favour of honour killing also disagree with extreme attitudes such as obedience, traditionalism, and authority (Ilic, 2016). Therefore, no correlation with hostility and deterrence of honour killing is aligned with the previous literature. Furthermore, no correlation between deterrence of honour killing and hostility/intolerance confirmed discriminant validity. Discriminant validity has provided sufficient evidence that two measures diverge from each other.

Henceforth, the construct validity of newly developed scale was supplemented by finding significant positive relationship between modern gender role attitudes and subscale of newly developed measure that is deterrents of honour killing, while significant negative relationship between modern gender role attitudes and Affirmation of honour killing. Overall, the results were in line with the hypotheses of the current study and results provide sufficient support for reliability, construct, and convergent validity for newly developed ATHKS comprised of two subscales.

CONCLUSION

The purpose of this study was to develop and validate scale measuring attitude towards honour killing. The present study established construct validity of the newly developed ATHKS. The scale can be used to measure attitude towards honour killing particularly in Pakistan and in general on Muslim population. The scale can be applied to individuals over the age of 18years and minimum education of matriculation. The new indigenously developed Scale has been proved to be reliable and valid.

LIMITATIONS, SUGGESTIONS, AND PRACTICAL IMPLICATIONS

The present study is not free of limitations and thus they have to be mentioned along suggestion to cater those in future research. The sample in the present study lacks generalizability in the sense that the data had only been collected from Islamabad and few cities of Punjab. Moreover, the items of the Scale comprised of religious and cultural norms and laws. Thereby, limiting the use of scale to Pakistan in particular and Muslim community in general. Future studies can be planned to employ data from other parts of Pakistan as honour killing is widely witnessed phenomenon.

Despite of the limitations, a newly constructed scale fulfilled the requirements of reliability and validity. However, addition of modification indices could have improved overall goodness-of-fit indexes. Future studies could employ few other measures to establish discriminant validity. By the help of this newly developed scale, it is easier to measure attitudes of people towards honour killing from multidimensional references. There was a dearth of likert-type scales for measuring attitude towards honour killing globally. Growing number of statistics in Pakistan on honour killing incidents motivated to develop a measure with reference to Pakistani context.

REFERENCES

1. Ali, R. (2001). The dark side of ‘Honour’.Shirkat Ga.
2. Anila, & Ansari, Z. A. (1992). Development of sex role attitude scale for Pakistan (Unpublished manuscript). National Institute of Psychology, Quaid-i-Azam University, Islamabad, Pakistan.
3. Babakus, E., & Mangold, W. G. (1992). Adapting the SERVQUAL scale to hospital services: An empirical investigation. Health Services Research, 26(6), 767.
4. Bowman, J. (2007). Honour: A history. New York, NY: Encounter Books.
5. Caffaro, F., Ferraris, F., & Schmidt, S. (2014). Gender differences in the perception of honour killing in individualist versus collectivistic cultures: Comparison between Italy and Turkey. Sex Roles, 71(9-10), 296-318.
6. Chesler, P. (2010). Worldwide trends in honour killings. Middle East Quarterly, 17(2), 3-11.
7. Cohan, J. A. (2009). Honour killings and the cultural defense. Cal. W. Int'l L J, 40, 177.
8. Cohan, J. A. (2010). Honour killings and the cultural defense. California Western International Law Journal, 40(2), 177-252.
9. Cohen, D., Nisbett, R. E., Bowdle, B. F., & Schwarz, N. (1996). Insult, aggression, and the southern culture of honour: An experimental ethnography. Journal of Personality and Social Psychology, 70(5), 945-960.
10. Cooney, M. (2014). Death by family: Honour violence as punishment. Punishment & Society, 16(4), 406-427.
11. Copp, J. E., Giordano, P. C., Manning, W. D., & Longmore, M. A. (2016). Couple‐level economic and career concerns and intimate partner violence in young adulthood. Journal of Marriage and Family, 78(3), 744-758.
12. Deol, S. S. (2014). Honour killings in India: A study of the Punjab State. International Research Journal of Social Sciences, 3(6), 7-16.
13. Dogan, R. (2011). Is honour killing a “Muslim phenomenon”? Textual interpretations and cultural representations. Journal of Muslim Minority Affairs, 31(3), 423-440.
14. Dogan, R. (2014). Different cultural understandings of honour that inspire killing: An inquiry into the defendant’s perspective. Homicide Studies, 18(4), 363-388.
15. Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX: Harcout Brace Jovanovich.
16. Eisner, M., & Ghuneim, L. (2013). Honour killing attitudes amongst adolescents in Amman, Jordan. Aggressive Behavior, 39(5), 405-417.
17. Elmquist, J., Shorey, R. C., Labrecque, L., Ninnemann, A., Zapor, H., Febres, J., & Stuart, G. L. (2016). The relationship between family-of-origin violence, hostility, and intimate partner violence in men arrested for domestic violence: Testing a mediational model. Violence Against Women, 22(10), 1243-1258.
18. Esqueda, C. W., & Harrison, L. A. (2005). The influence of gender role stereotypes, the woman's race, and level of provocation and resistance on domestic violence culpability attributions. Sex Roles, 53(11-12), 821-834.
19. Faqir, F. (2001). Intra family femicide in defence of honour: The case of Jordan. Third World Quarterly, 22(1), 65-82.
20. Field, A. (2009). Discovering Statistics Using SPSS: Introducing Statistical Method (3^rd ed.). Thousand Oaks, CA: Sage Publications.
21. Gilani, N., & Altaf, R. (2005).Tendencies of extremism among adolescents and post-adolescents in relation to parenting style. Pakistan Journal of Social and Clinical Psychology, 3(1-2), 27-40.
22. Golge, Z. B., Yavuz, M. F., Müderrisoglu, S., & Yavuz, M. S. (2003).Turkish university students’ attitudes toward rape. Sex Roles, 49, 653-661.
23. Gracia, E., Lila, M., & European Commission. (2015). Attitudes towards violence against women in the EU. Luxembourg: Publications Office of the European Union.
24. Greiff, S. (2010). No Justice in Justifications: Violence against women in the name of culture, religion, and tradition. Resource Paper, Global Campaign to Stop Killing and Stoning Women, 1-44.
25. Grzyb, M. A. (2016). An explanation of honour-related killings of women in Europe through Bourdieu’s concept of symbolic violence and masculine domination. Current Sociology, 64(7), 1036-1053.
26. Gudmundsson, E. (2009). Guidelines for translating and adapting psychological instruments. Nordic Psychology, 61(2), 29-45.
27. Guerra, V. M., Gouveia, V. V., de CR Araújo, R., de Andrade, J. M., & Gaudêncio, C. A. (2013). Honour Scale: Evidence on construct validity. Journal of Applied Social Psychology, 43(6), 1273-1280.
28. Gupte, M. (2013). The concept of honour: Caste ideology and patriarchy in rural Maharashtra. Economic and Political Weekly, 48(18), 72-81.
29. Gürsoy, E., & Arslan-Özkan, H. (2014). Turkish youth’s perception of sexuality/honour in relation to women. Journal of Psychiatric Nursing, 5(3), 149-159.
30. Hair, J., Jr., Anderson, R., Tatham, R., & Black, W. (1998). Multivariate data analysis (5^th ed.). Upper Saddel River, NJ: Prentice Hall.
31. Henson, R. K., & Roberts, J. K. (2006). Use of exploratory factor analysis in published research: Common errors and some comment on improved practice. Educational and Psychological Measurement, 66(3), 393-416.
32. Hill, S., & Marshall, T. C. (2018). Beliefs about sexual assault in India and Britain are explained by attitudes toward women and hostile sexism. Sex Roles, 79(7-8), 421-430.
33. Hillier, L., & Foddy, M. (1993). The role of observer attitudes in judgments of blame in cases of wife assault. Sex Roles, 29(9-10), 629-644.
34. Huda, S., & Kamal, A. (2017). Professionals perspective in portrayal of honour killings in Pakistani media. Pakistan Journal of Criminology, 9(1), 15-28.
35. Ilic, P. (2016). Honour killing attitudes among San Jose State University students. Themis: Research Journal of Justice Studies and Forensic Science, 4(1), 8.
36. Işık, R., & Sakallı Uğurlu, N. (2009).The development of Attitudes toward Honour Scale and Attitudes toward Violence against Women for Protecting Honour Scale with a student sample. Turkish Psychological Articles, 12(24), 25-27.
37. Jafri, A. H. (2008). Honour killing: Dilemma, ritual, understanding. Oxford University Press.
38. Jamali, H., & Shah, N. A. (2015). The origin of honour killing (karo-kari) in Sindh, Pakistan: A discussion on Islamic context. The Women-Annual Research Journal of Gender Studies, 7, 60-82.
39. Kamal, A., & Saqib, T. (2004). Exploring change in gender role attitudes (Unpublished. M.Phil Dissertation). National Institute of Psychology, Quaid-i-Azam, University, Islamabad, Pakistan
40. Khan, R. (2018). Attitudes towards ‘honour’ violence and killings in collectivist cultures: Gender differences in Middle Eastern, North African, South Asian (MENASA) and Turkish populations In J. L. Ireland, P, Birch, & C. A. Ireland (Eds.), International Handbook in Aggression: Current Issues and Perspectives (pp. 216-226). London: Routledge.
41. Linton, C., & Lorde, T. (2011). The influence of gender and gender role attitudes on domestic violence culpability assignment in Barbados.
42. Nunnally, J. C. (1994). Psychometric theory (3^rd ed.). Tata McGraw-Hill Education.
43. Pasquali, L. (2003). Theory of testing in psychology and education. Petrópolis, RJ: Vozes.
44. Pitt-Rivers, J. (1966). Honour and social status. In J. Peristiany (Ed.), Honour and shame: The values of Mediterranean society (pp. 19-77). London: Weidenfeld & Nicholson.
45. Rahim, S., Jahangir, F., & Holden, G. W. (2016). New instrument to assess attitudes toward honour killing: The relation between gender and education. FWU Journal of Social Sciences, 10(2), 132-136
46. Rahim, S., Jahangir, F., & Zeb, R. (2018). Development and validation of Attitude towards Honour Killing Scale. Pakistan Journal of Psychological Research, 33(1) 191-202.
47. Reddy, R. (2008). Gender, culture and the law: Approaches to ‘honour crimes’ in the UK. Feminist Legal Studies, 16(3), 305-321.
48. Rosenblatt, P. C. (2010). Shame and death in cultural context. In J. Kauffman (Ed.). The Shame of Death, Grief, and Trauma (pp. 113-137).East Sussex, London: Routledge.
49. Sakallı Uğurlu, N., & Akbaş, G. (2013). Honour and honour violence against women in honour cultures: Social psychological explanations. Türk Psikoloji Yazıları, 16(32), 76-91.
50. Schmitt, T. A. (2011). Current methodological considerations in exploratory and confirmatory factor analysis. Journal of Psycho-educational Assessment, 29(4), 304-321.
51. Sev'er, A., & Yurdakul, G. (2001). Culture of honour, culture of change: A feminist analysis of honour killings in rural Turkey. Violence Against Women, 7(9), 964-998.
52. Shaikh, M. A., Shaikh, I. A., Kamal, A., & Masood, S. (2010). Attitudes about honour killing among men and women: Perspective from Islamabad. Journal of Ayub Medical College Abbottabad, 22(3), 38-41.
53. Shaikh, M. A., Kamal, A., & Naqvi, I. (2015). Opinions of university students on honour killings: Perspective from Islamabad and Rawalpindi. ournal of the Pakistan Medical Association, 65, 421-424.
54. Shier, A., & Shor, E. (2016). “Shades of foreign evil”, “honour killings”, and “family murders” in the Canadian Press. Violence Against Women, 22(10), 1163-1188.
55. Stevens, J. (1996). Confirmatory and exploratory factor analysis, In J. P. Stevens (Ed.), Applied Multivariate Statistics for the Social Sciences (3^rd ed.), 362-427. Mahwah: Lawrence Erlbaum.
56. Vandello, J. A., & Cohen, D. (2003). Male honour and female fidelity: Implicit cultural scripts that perpetuate domestic violence. Journal of Personality and Social Psychology, 84(5), 997.
57. Waltermaurer, E. (2012). Public justification of intimate partner violence: A review of the literature. Trauma, Violence, & Abuse, 13(3), 167-175.
58. Warraich, S. A. (2005). ‘Honor killing’ and the law in Pakistan. In: Welchman, L., & Hossain, S. (Eds.) ‘Honor’: Crimes, Paradigms and Violence against Women, London: Zed Books.
59. Welchman, L., & Hossain, S. (2005). Introduction: Honour, rights, and wrongs. In Hossain, S. and Welchman, L. (Eds.), Honour: Crimes, paradigms and violence against women. London: Zed Books.
60. Welden, B. A. (2010). Restoring lost honour: Retrieving face and identity, removing shame, and controlling the familial cultural environment through ‘honour’ murder. Journal of Alternative Perspectives in the Social Sciences, 2(1), 380-398.
61. Wikan, U. (2008). In honour of Fadime: Murder and shame. University of Chicago Press.
62. Williams, B., Onsman, A., & Brown, T. (2010). Exploratory factor analysis: A five-step guide for novices. Australasian Journal of Paramedicine, 8(3), 1-13.
63. Zia Ullah, M. (2010). Honour killings in Pakistan: Under theoretical, legal and religious perspectives. Published master thesis, Malmo University, Sweden.