Validity and Reliability of the Malay Version of WHO Women’s Health and Life Experiences Questionnaire

Abstract

The Women’s Health and Life Experiences questionnaire measures the prevalence, health implications, and risk factors for domestic violence. This cross-sectional study was conducted to determine the validity and reliability of the Malay version of World Health Organization (WHO) Women’s Health and Life Experiences Questionnaire. Construct validity and reliability assessment of the Malay version of the questionnaire was done on 20 specific items that measure four types of intimate partner violence (IPV) act; controlling behaviors (CB), emotional violence (EV), physical violence (PV), and sexual violence (SV), which were considered as the domains of interest. Face-to-face interviewing method was used for data collection. A total of 922 women completed the interviews. The results showed that exploratory factor analysis of four factors with eigenvalues above 1 accounted for 63.83% of the variance. Exploratory factor analysis revealed that all items loaded above 0.40 and the majority of items loaded on factors that were generally consistent with the proposed construct. The internal consistency reliability was good. The Cronbach’s α values ranged from 0.767 to 0.858 across domains. The Malay version of WHO Women’s Health and Life Experiences Questionnaire is a valid and reliable measure of women’s health and experiences of IPV in Malaysia.

Keywords

validity reliability Malay Women’s Health and Life Experiences Questionnaire

Introduction

Intimate partner violence (IPV) refers to any act of violence that occurs between two people in a close relationship who share their personal and social lives intimately and are bound either by marital contract, cultural practices, or personal agreements. The nature of violent acts can be physical, sexual, emotional, or involve display of controlling behaviors, which can also be interpreted as a form of social violence (World Health Organization, 2002). The prevalence, and severity, of IPV against women is overwhelming as many women suffered from more than one type of violence in the relationship (Thompson, Basile, Hertz, & Sitterle, 2006; Yoshihama & Sorenson, 1994). Findings from a large population-based study by the World Health Organization (WHO) revealed that 15% to 71% of women had been physically and/or sexually abused by their partners at some point in their lives (Garcia-Moreno et al., 2006). While most women can easily identify physical and sexual violence, experience of emotional abuse or being controlled by their partners may not be readily discernible unless specified. Emotional or psychological abuse includes the use of words or gestures to humiliate, intimidate, or communicate the intent to cause physical harm, and controlling behaviors include acts to constrain the partner’s mobility as well as displays of irrational anger and extreme jealousy toward the partner (World Health Organization, 2002).

A range of questionnaires are available to investigate the prevalence of IPV. These include IPV screening tools such as the Hurt, Insulted, Threatened, or Screamed at instrument (Sherin, Sinacore, Li, Zitter, & Shakil, 1998), the Woman Abuse Screening Tool (Brown, Lent, Schmidt, & Sas, 2000), the Abuse Assessment Screen (Norton, Peipert, Zierler, Lima, & Hume, 1995), and the Partner Violence Screen (Feldhaus et al., 1997). Although these brief tools are useful to assist with diagnosis in clinical settings they have several shortcomings. A systematic review of 33 IPV screening tools showed that none of the instruments had sound psychometric properties and recommended further testing in varied settings and languages (Rabin, Jennings, Campbell, & Bair-Merritt, 2009). Some questionnaires have also been criticized for not being comprehensive. For example, the Hurt, Insulted, Threatened, or Screamed at instrument did not ask about sexual violence and the Partner Violence Screen only has questions about physical violence (Rabin et al., 2009).

In addition to the brief screening tools, there are also instruments intended for research (Thompson et al., 2006). Of available scales, the Revised Conflict Tactics Scales is perhaps the most widely used instrument for measuring IPV (Straus & Douglas, 2004). Nevertheless, it has also been criticized for not adequately measuring the context, severity, meanings, and motives of violence (DeKeseredy & Schwartz, 1998; Dobash & Dobash, 2004). Quite recently, the World Health Organization has developed a questionnaire to measure women’s health and IPV experiences (Garcia-Moreno et al., 2006). This questionnaire, the Women’s Health and Life Experiences questionnaire, was developed for the World Health Organization Multi-country Study on Women’s Health and Domestic Violence against Women (WHO VAW Study), a large population-based survey conducted on 24,097 women from 15 sites in 10 study countries (Bangladesh, Brazil, Ethiopia, Japan, Namibia, Peru, Samoa, Serbia and Montenegro, Thailand, and the United Republic of Tanzania) between year 2000 and 2003 (Garcia-Moreno et al., 2006). The Women’s Health and Life Experiences questionnaire has since became an internationally recognized tool with several advantages compared to other instruments. Besides broadly conceptualizing IPV with use of a range of behavior-specific questions to measure the prevalence, causes, frequency, and consequences of various types of violence, the questionnaire also assesses health implications of violence, identifies the risk and protective factors for violence, and explores strategies taken by the women who experienced violence. Findings of the WHO VAW study demonstrated the usefulness of the Women’s Health and Life Experiences questionnaire to explore the extent and characteristics of IPV (Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts, 2005). The original English version of the questionnaire has been translated and used for other population-based studies (Clark, Bloom, Hill, & Silverman, 2009; Valladares, Peña, Persson, & Högberg, 2005; Vung, Ostergren, & Krantz, 2008).

Data on IPV are relatively scarce in Malaysia. This lack of sound information has actually hindered the development of a national action plan for prevention of violence in this country (World Health Organization, 2006). Within the context of IPV, characteristics of perpetrators and victims, factors that provoke violent episodes, influence of social and environmental conditions and circumstances that make certain individuals prone to become either the perpetrator or the victim are among the important information needed to prevent and control the problem. Large population-based studies are thus warranted to fill these gaps with use of comprehensive standardized questionnaire. In this cross-sectional study, we aimed to translate and validate the Women’s Health and Life Experiences questionnaire into Malay language for measuring IPV among women in Malaysia.

Method

Introduction to the Study Area

Johor is the southernmost state of the Peninsular Malaysia. The state of Johor is divided into the districts of Johor Bahru, Kulai Jaya, Pontian, Kota Tinggi, Kluang, Segamat, Muar, Ledang, Batu Pahat, and Mersing. Johor Bahru District is the most densely populated metropolitan area in Johor with high concentration of industries while other districts of the state have a balanced mixture of agricultural activities and rural sectors as well as urban developments. In 2010, the population of Johor was estimated at 3.3 million with the majority (44.3%) of them lived in the Johor Bahru District (Department of Statistics Malaysia, 2010). The reported male to female ratio in Johor was 107:100. The selection of Johor as the study area was based on the common knowledge that the standard Malay language in Malaysia is spoken by the people of the state. In fact, the original source of the standard Malay language is said to be from the Riau-Johor dialect (Sariyan, 2002).

Study Design and Sample

A cross-sectional study was conducted at 10 sites selected from three districts of Johor, namely Johor Bahru, Pontian, and Kota Tinggi, between April 1, 2008 and March 31, 2009. The sites were purposely selected to represent women living in both rural and urban areas within the districts. The sites’ selection was based on the location of public health centers that provide health-care services under the Ministry of Health Malaysia (MOH) health-care system. Five urban health centers in Johor Bahru district and five rural health centers in Pontian and Kota Tinggi were chosen. The sampling frame consisted of women who attended the health centers during the data collection period. The inclusion criteria for the study were Malay women aged 18 and above, who were able to read and write fluently in Malay language. The participants had to be Malaysian and women who showed clear cognitive challenges were excluded. Convenient sampling was applied for selection of study subjects to get the prior determined sample size of 930. Ethical approval to conduct the study was obtained from the Human Research Ethics Committee, Universiti Sains Malaysia. Ethical approval was also sought from the Medical Research and Ethics Committee, MOH Malaysia since the study was conducted in the MOH premises.

The Instrument

We used the 10th version of the Women’s Health and Life Experiences questionnaire (World Health Organization, 2003). This questionnaire consisted of an administration form, a household selection form, a household questionnaire, a female questionnaire, and two reference sheets. The female questionnaire included 12 sections: (a) characteristics of the respondent and her community; (b) general health status; (c) reproductive health; (d) information regarding children; (e) characteristics of current or most recent partner; (f) attitudes toward gender roles; (g) experiences of partner violence; (h) injuries due to violence; (i) impact and coping mechanisms used by women who experience violence; (j) experiences of non-partner violence; (k) financial autonomy; and (l) respondent feedback to complete the interview.

The Translation Process

The translation process began with forward translation of the original core questionnaire from English to Malay language; done independently by two individuals whose native language was Malay and who were fluent in English. The two translations were then compared and revised to produce a forward-translation of the Women’s Health and Life Experiences questionnaire. This revision process was done by the researchers with help of an expert panel that consisted of the translators, one health professional with experience in managing domestic violence victims, and one researcher with experience in questionnaire translation and development. The aim of this process was to identify and resolve inadequate expressions or concepts in the translated versions, as well as to detect any discrepancies between the forward-translation and the original version of the questionnaire. These have resulted in changes of some words or expressions.

Later, the provisional Malay version of the questionnaire was sent to another two individuals who translated it back into English. The translators, who were blinded from the original questionnaire, could speak and write good English and Malay. The back-translation was also done independently. As in the initial translation, the back-translated versions were also compared and evaluated against the original questionnaire. This was to ensure that corresponding items in both versions convey similar meanings. Although ideally forms of language should be similar too, some variations were acknowledged on condition that the equivalence of meaning was guaranteed. The previous panel members were maintained while the new translators were included for the work to produce the preliminary Malay version of Women’s Health and Life Experiences questionnaire.

Cultural Adaptation, Validity, and Reliability of the Malay Version

Adaptation of the Women’s Health and Life Experiences questionnaire included reviewing the equivalences between the translated and the original version of the questionnaire as described by Meadows (2003): (a) conceptual equivalence to assess the relevance of domestic violence issues in the Malay culture; (b) item/content equivalence to assess whether the situations described or experiences evoked in the original version are applicable in the Malay culture; (b) semantic equivalence to assess whether the meaning of each item, word, or expression was maintained after translation into the Malay language; (d) operational equivalence to assess whether the methods of data collection will affect the results differently; (e) measurement equivalence to assess whether the psychometric properties of the Malay translation were similar to that of the original version; and (d) functional equivalence to assess achievements of the preceding equivalences.

The Women’s Health and Life Experiences questionnaire was developed not only to explore experiences of violence, but also to capture other health and health-related experiences in the women’s life. As such, this multidimensional questionnaire was designed and structured with complex skip pattern where probing questions are asked only at relevant respondents. This design is useful in long questionnaires such as the Women’s Health and Life Experiences questionnaire which has over 1,000 items in total. However, the employment of this design where respondents may skip or answer questions or sections based upon their previous responses has made it very complex to statistically assess the questionnaire validity as a whole. Therefore, based on the current and anticipated uses of the questionnaire, we decided that the construct validity of the Malay version of Women’s Health and Life Experiences questionnaire would only be done on 20 specific items (Box 1) that measure controlling behaviors (CB—7 items) and other three types of violence act, emotional violence (EV—4 items), physical violence (PV—6 items), and sexual violence (SV—3 items).

Box 1.

Items measuring violence experiences included in reliability and validity analysis of the Malay version of the Women’s Health and Life Experiences questionnaire.

Controlling behaviors

Tried to keep you from seeing your friends

Tried to restrict contact with your family of birth

Insisted on knowing where you are at all times

Ignored you and treated you indifferently

Got angry if you spoke with another man

Was often suspicious that you are unfaithful

Expected you to ask his permission before seeking health care for yourself

Emotional violence (EV)

Insulted you or made you feel bad about yourself

Belittled or humiliated you in front of other people

Done things to scare or intimidate you on purpose (e.g., by the way he looked at you, by yelling and smashing things)

Threatened to hurt you or someone you care about

Physical violence (PV)

Slapped you or thrown something at you that could hurt you

Pushed you or shoved you or pulled your hair

Hit you with his fist or with something else that could hurt you

Kicked you, dragged you, or beaten you up

Choked or burnt you on purpose

Threatened to use or actually used a gun, knife, or other weapon against you

Sexual violence (SV)

Physically forced you to have sexual intercourse when you did not want to

Ever had sexual intercourse you did not want to because you were afraid of what your partner or any other partner might do

Ever forced you to do something sexual that you found degrading or humiliating

All ever-partnered women, defined as women who ever had an intimate male partner (Garcia-Moreno et al., 2006), were asked to report their experience of CB, EV, PV, and SV. In particular, questions that explore partner’s CB asked the respondents if they thought the given situations were true for them. The response options offered were “yes,” “no,” and “don’t know.” In contrast, questions on EV, PV, and SV refer to actions that may have been perpetrated by the partner, and the response options given were “yes” and “no.” In addition, for each reported act of EV, PV, and SV, there were further questions about whether the incident happened during the past 12 months or before the past 12 months, and with what frequency (one, few, or many). However, in this study to assess the psychometric properties of the Malay version of Women’s Health and Life Experiences questionnaire, the timing and frequency of EV, PV, and SV were not included in the analyses.

Researchers seem to disagree whether middle responses like “don’t know” should be offered in questionnaires, particular in attitudinal questions, although many authors advocated the use of “don’t know” responses for factual questions because there is a possibility that respondents may not know the answer to a question (Groothuis & Whitehead, 2002). In IPV, the CB situations are often difficult to recognize, unlike the acts EV, PV, and SV. Most CB situations are covert and disguised by actions that appear to be normal. Unlike the perpetrator of CB, the victim may not realize or may choose to ignore the incident until it leads to certain negative consequence that necessitates her to recall the experience she had undergone and named it as CB. For example, women may remember clearly if they have been hit by their partner or if their partner had humiliated them in front of other people or physically forced them to have sexual intercourse. However, it may be difficult for the women to give definitive “yes” or “no” answer to whether they thought their partner generally gets angry if they speak to another man as it can be easily mistaken for the partner having a bad day or just his personality or he was being loving and caring. As such, the “don’t know” responses were considered as valid responses for assessing CB situations in the Women’s Health and Life Experiences questionnaire.

Training of Interviewers

The interviewers for this study consisted of 20 volunteers among the staff nurses working at the respective health centers. All the nurses were Malay females. The selection of interviewers, two from each center, was done based on the criteria described for the WHO VAW Study (Jansen, Watts, Ellsberg, Heise, & García-Moreno, 2004). Owing to the nature of their work, the nurses had some awareness of the issues of domestic violence and were adept in interviewing skills. The training of interviewers provides general information on the issues of IPV as well as focusing on familiarizing the nurses with the questionnaire and its protocol as well as standardization of assessments.

Pretest of the Malay Version

The preliminary Malay version of Women’s Health and Life Experiences questionnaire was pretested to ensure its clarity and comprehensiveness on conveniently selected 50 women—five from each health center. Respondents’ comments and feedbacks regarding difficulties in understanding and responding to the items were obtained. Appropriate changes were made accordingly to produce the final Malay version of Women’s Health and Life Experiences questionnaire.

Data Collection

A total of 930 women participated in this study. Face-to-face interviewing was used for the data collection method. The guidelines for research on domestic violence produced by the World Health Organization (2001) were strictly adhered to. All interviews were conducted in complete privacy, and the women were allowed to choose the time and place of interview to be conducted either at the respective health centers during the particular visit or later at the women’s home. Written informed consent was obtained from all participants and it was made clear to them about the possibility to withdraw at any point during the interview. The informed consent form was written in Malay language and the copy was approved for use in this study by the Human Research Ethics Committee, Universiti Sains Malaysia.

Data Analysis

Data entries and analyses of results were done using the IBM SPSS Statistics for Windows (version 20.0, IBM Corp., Armonk, NY), a statistical software package. Although there were some missing values scattered randomly through the data matrix, it involved less than 5% of the data point and thus considered nonproblematic (Tabachnick & Fidell, 2007). Answers for all 20 items that measure IPV in this study were complete. Characteristics of the participants were determined; categorical and numerical data were presented by frequencies and means respectively. The Kaiser–Meyer–Olkin Measure of Sampling and the Bartlett’s Test of Sphericity were performed to determine suitability of data for factor analysis. It was hypothesized that four factors would be obtained. Assuming the factors were correlated, first we performed oblique rotation by direct oblimin method. However, the resulting correlation matrix indicated that there was not enough variance to warrant oblique rotation (Tabachnick & Fidell, 2007). The highest correlation was 0.32 and it was limited to only one pair of factors. The remaining correlations were far below the threshold of 0.32 recommended by Tabachnick and Fidell (2007). Orthogonal rotation by varimax method was thus chosen for factor analysis. The cut point for loading to be interpreted was decided at 0.40. The internal consistency reliability of the factors was evaluated using the Cronbach’s α.

Results

Content Validity

The expert panel reviewed the Malay version of Women’s Health and Life Experiences questionnaire and unanimously agreed that all items were appropriate and adequately covered the constructs measured. However, the experts suggested some minor modifications to address issues relevant to the local setting. The revisions include removal of the household selection form, which was not applicable in this study as the women in this case were a conveniently selected clinic-based sample. Other adaptations involved inserting or replacing responses to some specific items in section 9 and section 10 to appropriately suit the local scenario, particularly responses concerning religious leaders. The response lists for the items were expanded and rearranged to comprise main local religious leaders with whom the respondents are more familiar. As Islam is the official religion of Malaysia and Muslims form the majority of Malaysian population, instead of having priest at the start of the list, Imam—the title given for Muslim religious leader who leads prayers in mosques—was placed first. Careful attention was given to ensure that the questionnaire does not lose its psychometric properties.

Pretest in Target Population

The preliminary Malay version of Women’s Health and Life Experiences questionnaire was pretested in a sample of 50 Malay women. One woman did not complete the interview, leaving a total of 49 respondents. Their mean age was 32.4 (SD = 7.13) years and the majority were currently married (93.3%). In general, the respondents agreed they had clearly understood the questionnaire. Only minor formatting problems were noted and changes were done accordingly.

Characteristics of the Respondents

Of 930 women who initially participated, eight did not complete the interview, leaving a total of 922 respondents. All women who withdrew had their interviews conducted at the health centers and the main reason given for withdrawing was time constraint. The women also refused the interviewers’ offer for a later meeting at their home. Since participation in this study was purely voluntary and the women have the right to withdraw at any point during the interview, no further efforts were made to probe on the reason given or to pursue the interview.

All 922 respondents were Malay Muslim. Their sociodemographic characteristics are shown in Table 1. At the time of interview, 870 women were married (94.4%). Their mean age was 31.0 years (SD = 6.85) that ranged from 18 to 54 years. Lifetime experiences of violence reported by 885 ever-partnered women are shown in Table 2. The prevalence of any form of physical and sexual violence among the women were 47 (5.3%) and 22 (2.5%) respectively. Further analysis showed that none of the respondents answered “yes” to all violence items while 43.6% of the respondents answered “no” to all items and one respondent (0.1%) answered “don’t know” to all 7 CB items. The mean duration of the interview was 60 min (SD = 24.04) with most of the sessions lasted less than 1 hr (47.1%) and between 1 to 2 hr (52.0%). Only eight interviews lasted more than 2 hr.

Table 1.

Sociodemographic Characteristics of Respondents (n = 922).

Characteristics	Frequency (%)
Age (years)
≤ 19	20 (2.2)
20-24	156 (16.9)
25-29	254 (27.5)
30-34	200 (21.7)
35-39	168 (18.2)
40-44	104 (11.3)
45-49	18 (2.0)
≥ 50	2 (0.2)
Education level
None	4 (0.4)
Primary	80 (8.7)
Secondary	732 (79.4)
Higher	106 (11.5)
Marital status
Currently married	870 (94.4)
Living with man, not married	2 (0.2)
Having regular sexual partner, but living apart	2 (0.2)
Not currently married or involved in sexual relationship	48 (5.2)
Income earner
Yes	485 (56.2)
No	437 (47.4)

Table 2.

Lifetime Experiences of Violence Reported by Respondents (n = 885).

Characteristics	Frequency (%)
Physical violence (PV)
Slapped	39 (4.4)
Pushed or shoved	22 (2.5)
Hit with fist	21 (2.4)
Kicked, dragged or beaten	18 (2.0)
Choked or burnt	6 (0.7)
Threatened to use or used a weapon	9 (1.0)
Controlling behavior (CB)
Keep from seeing friends	58 (6.6)
Restrict contact with family	17 (1.9)
Insisted on knowing	123 (13.9)
Ignored and treated indifferently	24 (2.7)
Got angry	108 (12.2)
Suspicious	53 (6.0)
Ask permission	394 (445)
Emotional violence (EV)
Insulted	65 (7.3)
Belittled	43 (4.9)
Intimidated	53 (6.0)
Threatened to hurt	29 (3.3)
Sexual violence (SV)
Forced to have sexual intercourse	12 (1.4)
Afraid what partner might do	21 (2.4)
Did something degrading	5 (0.6)

Construct Validity

Twenty items were analyzed. The Kaiser-Meyer-Olkin value was high at 0.864 and the Bartlett’s Test of Sphericity was significant (p < .001), indicating a satisfactory factor analysis. Factor analysis of four factors with eigenvalues above 1 accounted for 63.83% of the variance with the first factor accounted for 33.25% of the variance, the second 17%, the third 7.56%, and the fourth 6.02%.

Loadings of items on factors and communality values of items are shown in Table 3. The communality values of most items were above 0.5 which indicate that the solution explains most of the variation for those items. Four items, “choked or burnt,” “threatened to use or used weapons,” “got angry,” and “ask permission” did not fit well with the factor solution with communalities less than 0.50. However, the items were retained because they loaded significantly on one factor. With a 0.40 cut point for inclusion of an item in interpretation of factors, all 20 items loaded higher than the set threshold. We also compared our factor loadings with the criteria established by Comrey and Lee (1992) that loadings in excess of 0.71 are considered excellent, 0.63 are very good, 0.55 are good, 0.45 are fair, and 0.32 to be poor. Results of our factor analysis presented in Table 3 showed that Factor 1 included 5 of 6 PV items (hit with fist, kicked, dragged or beaten, pushed or shoved, choked or burnt, threatened to use or used a weapon) and 1 EV item (threatened to hurt). All CB items loaded on Factor 2 and all SV items loaded excellently on Factor 4 without additional items. Factor 3 appeared to have fairly high loadings from 3 EV items (insulted, belittled, and intimidated) and 1 PV item (slapped). Besides loading on Factor 3, “slapped” also loaded on Factor 2. Nevertheless, while its loading on Factor 3 can be considered good at 0.552, its loading on Factor 2 was only fair at 0.500. In general, the majority of items loaded together appropriately in accordance with the proposed construct. Labels for the factors thus follow the corresponding types of violence as shown in Table 3.

Table 3.

Communalities (h²) and Factor Loadings by Varimax Rotation With Kaiser Normalization (n = 885).

	Factor 1	Factor 2	Factor 3	Factor 4
Items	PV	CB	EV	SV	h2
Hit with fist	0.853				0.837
Kicked, dragged or beaten	0.845				0.810
Pushed or shoved	0.764				0.691
Threatened to hurt	0.695				0.711
Choked or burnt	0.688				0.478
Threatened to use or used weapon	0.437				0.372
Restrict contact with family		0.840			0.728
Ignored and treated indifferently		0.811			0.702
Keep from seeing friends		0.800			0.686
Suspicious		0.768			0.612
Insisted on knowing		0.747			0.577
Got angry		0.575			0.467
Ask permission		0.531			0.373
Insulted			0.797		0.734
Belittled			0.725		0.609
Intimidated			0.686		0.634
Slapped	0.500		0.552		0.614
Did something degrading				0.810	0.665
Forced to have sexual intercourse				0.801	0.729
Afraid what partner might do				0.766	0.736

Note: Values 0.40 and lower were suppressed. Proposed factor labels: PV = Physical Violence; CB = Controlling Behaviors; EV = Emotional Violence; and SV = Sexual Violence.

Internal Consistency Reliability

The Malay version of Women’s Health and Life Experiences questionnaire showed good internal consistency with Cronbach’s α values (n = 885) that ranged from 0.767 to 0.858 across four different measures of violence (Table 4). The Cronbach’s α value of the overall violence scale was 0.823.

Table 4.

Internal Consistency of the Malay Version of the Women’s Health and Life Experiences Questionnaire (n = 885).

Measure	No. of items	Cronbach’s α
Controlling behavior (CB)	7	0.807
Emotional violence (EV)	4	0.818
Physical violence (PV)	6	0.858
Sexual violence (SV)	3	0.767

Discussion

The Women’s Health and Life Experiences questionnaire was developed specifically for the WHO VAW Study that in general aimed to obtain sound and comparable data within and across culturally diverse communities on IPV. To gather this information from all 15 sites of the 10 study countries, the questionnaire has been translated and pretested into 14 relevant languages such that it was considered to be cross-culturally appropriate with good reliability and validity (Garcia-Moreno et al., 2005; Garcia-Moreno et al., 2006). However, we cannot assume the performance will remain satisfactory in our language, culture, or population. Thus, besides determining the psychometric properties, adaptation of the Women’s Health and Life Experiences questionnaire into Malay emphasized on achieving equivalences between the translated version and the original English version. This assessment was done following the translation process to ensure that the questionnaire performs equally well across different cultural groups.

Psychometric analysis was performed to establish the appropriateness of items used to assess the different types of violence act. Four types of violence dimensions—CB, PV, EV, and SV—were measured. Both validity and reliability of the question-items were determined. Validity refers to the accuracy of the questionnaire to measure what it aims to measure, while reliability refers to the consistency of the questionnaire in reproducing the values when applied in different situations (Carmines & Zeller, 1991; Saw & Ng, 2001). Results of our study suggested that the Malay version of Women’s Health and Life Experiences questionnaire is a valid and reliable instrument to determine the prevalence of IPV against women. The content-related validity of the Malay version of Women’s Health and Life Experiences questionnaire was obtained by having a panel of experts to systematically examine the questionnaire. The appropriateness of each item and overall coverage of the domains, whether the items correspond to the concept, construct, or domains being measured was reviewed. The panel concluded collectively that the questionnaire was content valid. The strength of relationship among items was good and the data met the criteria for factor analysis. Variance explained by the factors was 63.8%, supporting the content validity of the questionnaire. Item communalities were mostly high. Exploratory factor analysis also revealed high loadings of items on factors that were generally consistent with the proposed construct. Only two items, “slapped” and “threatened to hurt,” did not seem to fit the theory developed in previous research (Garcia-Moreno et al., 2005).

The item “threatened to hurt,” which was classified as an EV act, has a very good loading on Factor 1 that collected most PV items. This result suggested that our respondents viewed the threat of violence as comparable with the act itself, meaning that it is possible that their partners actually acted on their threats. It has been reported that emotional abuse often precedes or accompanies physical abuse in IPV (O’Leary, 1999; Tjaden & Thoennes, 2000). A follow-up study on a sample of 393 couples from the counties of New York demonstrated the progression of psychological to physical forms of violence (Murphy & O’Leary, 1989). Similarly, a study performed on 87 battered women at shelters in Norway showed that women who received threats by their partners were at higher risk of being physically abused, particularly when the threats were repeated (Alsaker, Kristoffersen, Moen, & Baste, 2011).

The item “slapped” loaded substantially on two factors—EV and PV. This indicates that the item was related to both factors. Complex variables like this can be handled in few ways. One option is to try different rotation methods. However, in our case, the option of an oblique rotation has already been ruled out. Another recommended solution is to drop “slapped” from the factor analysis (Costello & Osborne, 2005). However, its high communality value of 0.614 indicated that the item fits well with the solution, which explains most of its variation. In addition, the prevalence of women being slapped in violence episodes has also been reported to be high (Hassan et al., 2004; Saddki, Suhaimi, & Daud, 2010; Yoshihama & Sorenson, 1994). As such, we eventually decided to keep the item and treat it as belonging to the factor on which it has the highest loading.

Slap is considered a moderate PV because the chance of being injured due to slap is actually fairly low (World Health Organization, 2005). In agreement, a study of 242 hospital records of female IPV victims treated at a One-Stop-Crisis-Centre in Malaysia showed that most injuries were caused by punch and kick, rather than slap (Saddki et al., 2010). A study conducted at 10 emergency departments located in two Midwestern cities of the United States reported similar results where most of the women were injured by being punched or pushed, or kicked (Muelleman, Lenaghan, & Pakieser, 1996). Higher loading of “slapped” on EV factor as compared to PV factor suggested that the experience of being slapped affected the women’s emotional state more than her physical. The face is most exposed to public view. It represents a person’s identity and plays an important role in social interactions (Zebrowitz & Montepare, 2008). It is generally accepted that the face is the focus of attention when one’s physical beauty and attractiveness is appraised. It is thus possible that a slap was viewed by our respondents as a degrading act that undermined their self-esteem, leaving them emotionally defeated. Given that self-esteem has been linked to a range of emotional states, particularly self-relevant emotions such as feelings of pride and shame (Brown et al., 2000), it is thus plausible that the physical act of slapping was associated with EV more than PV by the women in our study.

Although the purpose of a slap may be to humiliate rather than to injure, a slap is an act of violence regardless of the circumstances and consequences. Being slapped by the partners (4.4%) was the most common form of PV experienced by the women in our study. This prevalence was lower than the results of the WHO VAW study that ranged from 9% in Japan to 52% in provincial Peru (World Health Organization, 2005). The vast difference in the prevalence of violence between and within settings is not easily explained, as the causes of IPV are complex; although, male superiority in relationship and existing culture of violence have been identified as the necessary factors in IPV occurrence (Jewkes, 2002). In addition, it is possible that reluctance to disclose IPV experiences might have contributed to the overall low rate of violence experienced by our sample. Reasons for not reporting violence may vary among individuals. Personal factors such as failure to recognize an incident as violence, acceptance of violence, hope for change, financial dependence, stigma and fear, as well as barriers related to health care system and criminal justice system were among the common reasons given by victims of IPV (Hegarty & Taft, 2001; Hien & Ruglass, 2009; Montalvo-Liendo, Wardell, Engebretson, & Reininger, 2009; Naved, Azim, Bhuiya, & Persson, 2006).

The Cronbach’s α for the total scale was high at 0.823 and the values for all domains were greater than 0.70, which is indicative of a reliable scale. Our results are comparable with that reported in the WHO VAW study, which was 0.66, 0.73, and 0.81 for the SV, CB, and PV respectively. It should be noted that the WHO VAW study omitted the EV domain from reliability analysis owing to the complexity in measuring EV and investigations of EV was considered a starting-point rather than a comprehensive measure (Garcia-Moreno et al., 2006). However, our results showed that the internal consistency of the EV domain was good with high Cronbach’s α of 0.818. Further, all four items in the EV domain loaded more than 0.5, suggesting an acceptable fit. It was also noted that the Cronbach’s α value for SV domain was the lowest in both our study and the WHO VAW study. This may be due to the respondents’ reluctance to recall, confide, or discuss experiences of SV in a face-to-face interview. Feedback from our interviewers corroborated the account. Poor correlation in sexual activity item was also reported by Saddki et al. (2009) in a study that utilized interviewer-administered approach to assess the psychometric properties of the Malay version of World Health Organization Quality of Life (WHOQOL-HIV BREF) among a sample of HIV positive patients in the state of Kelantan, Malaysia. Sensitivity in divulging sexual experiences has been observed and reported not only in our culture, but also in others (Babu & Kar, 2009; Gossaye et al., 2003; Sakthong, Schommer, Gross, Sakulbumrungsil, & Prasithsirikul, 2007).

In general, the results of our study demonstrated successful adaptation of the WHO Women’s Health and Life Experiences questionnaire into the Malay language. However, there are few limitations that need to be considered. The first limitation concerns the number of samples. There is a range of recommendations for determining sample size in factor analysis. Some recommendations were based on absolute number of subjects. For example, Gorsuch (1983) recommended a minimum sample size of 100 while Comrey and Lee (1992) suggested a higher minimum sample size of 500. Some other recommendations were based on minimum subject-to-item ratio such as 5:1 (Gorsuch, 1983) and 10:1 (Munro, 2001). In addition, MacCallum, Widaman, Zhang, and Hong, (1999) considered other factors such as communalities and degree of over-determination of factors in determining the minimum sample size. By most standards, our sample size of 922 was more than enough to reduce 20 violence acts into four types of violence using factor analysis. However, the number of sample in the subgroups of women experiencing violence was quite small such that it was not possible to perform between group comparisons. Second, the interviewers were recruited among the nurses who worked at the respective health centers, whom some respondents may have known quite well. Familiarity with interviewers was reported to be helpful in some instances where respondents were able to express themselves well but at the same time it seems to add apprehension to others in divulging personal intimate experiences which may lead to error in reporting. Similar remarks were made by Vung, Ostergren, and Krantz (2008) in an IPV study in rural Vietnam. Third, the sample consisted of women who visited the health centers for various health-related reasons but violence, thus our results may not be inferred to other victims of violence in the community.

Findings from our study demonstrated that the Malay version of Women’s Health and Life Experiences questionnaire is a valid and reliable measure of women’s health and experiences of IPV in Malaysia. The items were culturally appropriate and properly worded. Although it is a long questionnaire with complex design, the use of simple and direct language greatly improved the ease with which interviewers and respondents were able to complete it. Instructions on where the interviewers should probe for more information or how replies should be recorded are appropriately placed, and the use of a booklet also made it easier to progress through the questionnaire. In addition, the items and domains were arranged such that the interview starts with relatively easy questions on characteristics of the community and the general health status of the respondents before progressing to the more sensitive questions on experiences of violence. Further explanation and confidentiality assurance given at the start of such questions also give the respondents the sense of being in control of the interview where they do not have to answer the questions that they do not want to. At this critical stage, the interviewers were also required to ask permission from the respondents to continue with the interview process. In addition, use of explicit “refused/no answer” options for sensitive items may have helped to reduce item nonresponse in general. Thus, although there were some missing answers detected at random, answers for all 20 items that measure IPV in this study were essentially complete. These important features of the Women’s Health and Life Experiences questionnaire greatly facilitated our data collection and analysis and helped directly to achieve our research objectives.

Footnotes

Acknowledgements

We would like to express our sincere gratitude to the World Health Organization for the use of the Women’s Health and Life Experiences Questionnaire. Special thanks go to Dr. Henrica Jansen and Dr. Claudia Garcia-Moreno for the kind assistance. We also gratefully acknowledge the Ministry of Health Malaysia for the approval to conduct this study at the selected health centers (Pasir Gudang, Tampoi, Kempas, Majidee, Larkin, Air Tawar 2, Sening, Parit Ismail, Ayer Baloi, Serkat). We also thank all interviewers who assisted in data collection and the women who shared their experiences with us.

Authors’ Note

All authors contributed to the conception and design of the study. Zaharah Sulaiman, Siti Hawa Ali, Tengku Nur Fadzilah Tengku Hassan, and Zabedah Baharudin contributed to data acquisition and management. Norkhafizah Saddki, Zaharah Sulaiman, Siti Hawa Ali, Tg Nur Fadzilah Tg Hassan, Sarimah Abdullah, Azriani Ab Rahman, Tengku Alina Tengku Ismail, and Rohana Abdul Jalil contributed to data analyses and interpretation. Norkhafizah Saddki contributed to preparation of the manuscript. All authors revised and approved the final manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was funded by the Universiti Sains Malaysia short term grant (304/PPSP/6131582).

Author Biographies

Norkhafizah Saddki is a senior lecturer at the School of Dental Sciences, Universiti Sains Malaysia. She received her Master of Community Medicine (Oral Health) from similar institution. Her research focuses on women’s oral health issues, in particular among antenatal mothers. She’s also interested in other health issues like HIV/AIDS and intimate partner violence.

Zaharah Sulaiman is a senior lecturer at the Women’s Health Development Unit, School of Medical Sciences, Universiti Sains Malaysia. She is currently doing her PhD in breastfeeding among working women at the La Trobe University, Victoria, Australia.

Siti Hawa Ali is an associate professor at the School of Health Sciences, Universiti Sains Malaysia. She is a trained social worker graduated with MA and CQSW from the University of Kent at Canterbury. Currently her research focus is on violence against women particularly intimate partner violence as well as on child abuse. She is also actively promoting good practice in health sector in supporting people experiencing violence.

Tengku Nur Fadzilah Tengku Hassan received a BSc in Mathematics from Universiti Kebangsaan Malaysia. Currently she is a postgraduate student at the Women’s Health Development Unit, School of Medical Sciences, Universiti Sains Malaysia. Her research interest is in women’s health particularly in intimate partner violence issues. Her postgraduate research focuses on coping strategies among battered women in Malaysia.

Sarimah Abdullah is a senior lecturer at the Biostatistics and Research Methodology Unit, School of Medical Sciences, Universiti Sains Malaysia. She received Master of Community Medicine (Epidemiology and Biostatistics) from Universiti Sains Malaysia. Her research focuses on the time series modeling and forecasting on communicable and non-communicable diseases. She is also interested in research on domestic violence against women.

Azriani Ab Rahman is a medical lecturer at the School of Medical Sciences, Universiti Sains Malaysia. She received Master of Community Medicine, subspecialised in family health. Her research focuses on adolescent sexual and reproductive health. Her other research interests include health of adolescents and children with disability.

Tengku Alina Tengku Ismail is a senior lecturer in Family Health at the Department of Community Medicine, School of Medical Sciences, Universiti Sains Malaysia. She is currently doing her PhD in infant nutrition. Her research interests include breastfeeding, family planning and maternal health.

Rohana Abdul Jalil is currently a Senior Lecturer at the Department of Community Medicine, School of Medical Sciences, Universiti Sains Malaysia. Her research interest is in community health particularly among women.

Zabedah Baharudin is currently the Deputy Director, Medical Practice Division, Ministry of Health Malaysia. During the time of study, she was the Deputy Director of Health for the state of Johor. She received a master’s degree in public health.

References

Alsaker

Kristoffersen

Moen

B. E.

Baste

(2011). Threats and acts of intimate partner violence reported by users at Norwegian women’s shelters. Journal of Interpersonal Violence, 26, 950-970.

Babu

B. V.

Kar

S. K.

(2009). Domestic violence against women in eastern India: A population-based study on prevalence and related issues. BMC Public Health, 9, 129.

Brown

J. B.

Lent

Schmidt

Sas

(2000). Application of the Woman Abuse Screening Tool (WAST) and WAST-short in the family practice setting. Journal of Family Practice, 49, 896-903.

Carmines

E. G.

Zeller

R. A.

(1991). Reliability and validity assessment. Newbury Park, CA: Sage.

Clark

C. J.

Bloom

D. E.

Hill

A. G.

Silverman

J. G.

(2009). Prevalence estimate of intimate partner violence in Jordan. Eastern Mediterranean Health Journal, 15, 880-889.

Comrey

A. L.

Lee

H. B.

(1992). A first course in factor analysis (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Costello

A. B.

Osborne

J. W.

(2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7). Retrieved from http://pareonline.net/getvn.asp?v=10&n=7

DeKeseredy

W. S.

Schwartz

M. D.

(1998). Measuring the extent of woman abuse in intimate heterosexual relationships: A critique of the Conflict Tactics Scales. VAWnet Applied Research Forum. Retrieved from http://www.csaj.org/documents/178.pdf

Department of Statistics Malaysia (2010). Basic population characteristics by administrative districts. Kuala Lumpur, Malaysia: Author.

10.

Dobash

R. P.

Dobash

R. E.

(2004). Women’s violence to men in intimate relationship. Working on a puzzle. British Journal of Criminology, 44, 324-349.

11.

Feldhaus

K. M.

Koziol-McLain

Amsbury

H. L.

Norton

I. M.

Lowenstein

S. R.

Abbott

J. T.

(1997). Accuracy of 3 brief screening questions for detecting partner violence in the emergency department. Journal of the American Medical Association, 277, 1357-1361.

12.

Garcia-Moreno

Jansen

H. A. F. M.

Ellsberg

Heise

Watts

(2005). WHO multi-country study on women’s health and domestic violence against women. Initial results on prevalence, health outcomes and women’s responses. Geneva, Switzerland: World Health Organization.

13.

Garcia-Moreno

Jansen

H. A. F. M.

Ellsberg

Heise

Watts

C. H.

WHO Multi-country Study on Women’s Health and Domestic Violence against Women Study Team (2006). Prevalence of intimate partner violence: findings from the WHO multi-country study on women’s health and domestic violence. Lancet, 368(9543), 1260-1269.

14.

Gorsuch

R. L.

(1983). Factor analysis. Hillsdale, NJ: Lawrence Erlbaum.

15.

Gossaye

Deyessa

Berhane

Ellsberg

Emmelin

Ashenafi

Hogberg

(2003). Women’s health and life events study in rural Ethiopia. Ethiopian Journal of Health Development, 17(Second Special Issue), 1-49.

16.

Groothuis

P. A.

Whitehead

J. C.

(2002). Does don’t know mean no? Analysis of “don’t know” responses in contingent valuation questions. Applied Economics, 34, 1935-1940.

17.

Hassan

Sadowski

L. S.

Bangdiwala

S. I.

Vizcarra

Ramiro

De Paula

C. S.

Mitra

M. K.

(2004). Physical intimate partner violence in Chile, Egypt, India and the Philippines. Injury Control and Safety Promotion, 11(2), 111-116.

18.

Hegarty

K. L.

Taft

A. J.

(2001). Overcoming the barriers to disclosure and inquiry of partner abuse for women attending general practice. Australian and New Zealand Journal of Public Health, 25, 433-437.

19.

Hien

Ruglass

(2009). Interpersonal partner violence and women in the United States: An overview of prevalence rates, psychiatric correlates and consequences and barriers to help seeking. International Journal of Law Psychiatry, 32(1), 48-55.

20.

Jansen

H. A. F. M.

Watts

Ellsberg

Heise

García-Moreno

(2004). Interviewer training in the WHO multi-country study on women’s health and domestic violence. Violence Against Women, 10, 831-849.

21.

Jewkes

(2002). Intimate partner violence: Causes and prevention. Lancet, 359(9315), 1423-1429.

22.

MacCallum

R. C.

Widaman

K. F.

Zhang

Hong

(1999). Sample size in factor analysis. Psychological Methods, 4(1), 84-99.

23.

Meadows

K. A.

(2003). So you want to do research? 5: Questionnaire design. British Journal of Community Nursing, 8, 562-570.

24.

Montalvo-Liendo

Wardell

D. W.

Engebretson

Reininger

B. M.

(2009). Factors influencing disclosure of abuse by women of Mexican descent. Journal of Nursing Scholarship, 41, 359-367.

25.

Muelleman

R. L.

Lenaghan

P. A.

Pakieser

R. A.

(1996). Battered women: Injury locations and types. Annals of Emergency Medicine, 28, 486-492.

26.

Munro

B. H.

(2001). Statistical methods for health care research (4th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.

27.

Murphy

C. M.

O’Leary

K. D.

(1989). Psychological aggression predicts physical aggression in early marriage. Journal of Consulting and Clinical Psychology, 57, 579-582.

28.

Naved

R. T.

Azim

Bhuiya

Persson

L. A.

(2006). Physical violence by husbands: Magnitude, disclosure and help-seeking behavior of women in Bangladesh. Social Science and Medicine, 62, 2917-2929.

29.

Norton

L. B.

Peipert

J. F.

Zierler

Lima

Hume

(1995). Battering in pregnancy: An assessment of two screening methods. Obstetrics and Gynecology, 85, 321-325.

30.

O’Leary

K. D.

(1999). Psychological abuse: A variable deserving critical attention in domestic violence. Violence and Victims, 14, 3-23.

31.

Rabin

R. F.

Jennings

J. M.

Campbell

J. C.

Bair-Merritt

M. H.

(2009). Intimate partner violence screening tools: A systematic review. American Journal of Preventive Medicine, 36, 439-445.e434.

32.

Saddki

Noor

M. M.

Norbanee

T. H.

Rusli

M. A.

Norzila

Zaharah

Zarina

Z. A.

(2009). Validity and reliability of the Malay version of WHOQOL-HIV BREF in patients with HIV infection. AIDS Care, 21, 1271-1278.

33.

Saddki

Suhaimi

A. A.

Daud

(2010). Maxillofacial injuries associated with intimate partner violence in women. BMC Public Health, 10, 268.

34.

Sakthong

Schommer

J. C.

Gross

C. R.

Sakulbumrungsil

Prasithsirikul

(2007). Psychometric properties of WHOQOL-BREF-THAI in patients with HIV/AIDS. Journal of the Medical Association of Thailand, 90, 2449-2260.

35.

Sariyan

(2002). Prosiding Kongres Bahasa Dan Persuratan Melayu IV. Kuala Lumpur, Malaysia: Dewan Bahasa dan Pustaka.

36.

Saw

S. M.

T. P.

(2001). The design and assessment of questionnaires in clinical research. Singapore Medical Journal, 42(3), 131-135.

37.

Sherin

K. M.

Sinacore

J. M.

X. Q.

Zitter

R. E.

Shakil

(1998). HITS: A short domestic violence screening tool for use in a family practice setting. Family Medicine, 30, 508-512.

38.

Straus

M. A.

Douglas

E. M.

(2004). A short form of the Revised Conflict Tactics Scales, and typologies for severity and mutuality. Violence and Victims, 19, 507-520.

39.

Tabachnick

B. G.

Fidell

L. S.

(2007). Using multivariate analysis (5th ed.). Upper Saddle River, NJ: Pearson Allyn & Bacon.

40.

Thompson

M. P.

Basile

K. C.

Hertz

M. F.

Sitterle

(2006). Measuring intimate partner violence victimization and perpetration: a compendium of assessment tools. Atlanta, GA: Centers for Disease Control and Prevention, National Center for Injury Prevention and Control.

41.

Thompson

R. S.

Bonomi

A. E.

Anderson

Reid

R. J.

Dimer

J. A.

Carrell

Rivara

F. P.

(2006). Intimate partner violence: Prevalence, types, and chronicity in adult women. American Journal of Preventive Medicine, 30, 447-457.

42.

Tjaden

Thoennes

(2000). Extent, nature, and consequences of intimate partner violence: Findings from the National Violence Against Women Survey. Washington, DC: National Institute of Justice and the Centers for Disease Control and Prevention.

43.

Valladares

Peña

Persson

L. A.

Högberg

(2005). Violence against pregnant women: prevalence and characteristics. A population-based study in Nicaragua. BJOG: An International Journal of Obstetrics and Gynaecology, 112, 1243-1248.

44.

Vung

N. D.

Ostergren

P. O.

Krantz

(2008). Intimate partner violence against women in rural Vietnam—different socio-demographic factors are associated with different forms of violence: Need for new intervention guidelines? BMC Public Health, 8, 55.

45.

World Health Organization (2001). Putting women’s safety first: Ethical and safety recommendations for research on domestic violence against women. Geneva, Switzerland: Author.

46.

World Health Organization (2002). World report on violence and health: Summary. Geneva, Switzerland: Author.

47.

World Health Organization (2003). WHO multi-country study on women’s health and life experiences. Core questionnaire. Geneva, Switzerland: Author.

48.

World Health Organization (2005). WHO multi-country study on women’s health and domestic violence against women: Summary report of initial results on prevalence, health outcomes and women’s responses. Geneva, Switzerland: Author.

49.

World Health Organization (2006). National report on violence and health for Malaysia. Kobe, Japan: World Health Organization Centre for Health Development.

50.

Yoshihama

Sorenson

S. B.

(1994). Physical, sexual, and emotional abuse by male intimates: Experiences of women in Japan. Violence and Victims, 9, 63-77.

51.

Zebrowitz

L. A.

Montepare

J. M.

(2008). Social psychological face perception: why appearance matters. Social and Personality Psychology Compass, 2, 1497-1517.