Adapted tool for the assessment of domestic violence against women in a low-income country setting: a reliability analysis

Introduction

Violence against women (VAW) is a global public health pandemic and a serious human rights violation. Worldwide, one-in-three women has experienced VAW.^1–8 Domestic VAW is a common experience in the lives of women in Ethiopia. A World Health Organization’s (WHO’s) multi-country study indicated that domestic VAW in Ethiopia was 71%, which is the highest in the world.⁹ A systematic review conducted in Ethiopia (2000–2014) indicated that domestic VAW is a common phenomenon ranging from 20% to 78%.¹⁰ Women’s favorable attitude toward justifiable wife-beating, exacerbated by traditional gender-norms is a key underlying factor explaining domestic VAW. Currently, women’s receptive attitude toward justified wife-beating has declined from 81%¹¹ to 69%;¹² however, this is still unacceptably high.

In response to the high prevalence of domestic VAW, the government of Ethiopia has incorporated women’s right and gender equality in the constitution [Art-35 and 89(7)¹³ and other proclamations: Criminal code under proclamation No 414/2004 (Art 564)¹⁴ and Family Code Proclamation No 213/2000].¹⁵ Violence against a marriage partner or a person cohabiting, even in an irregular union, is prohibited. Moreover to help implement this, the Ethiopian Ministry of Health has developed a standard operating procedure for the response and prevention of VAW in 2016.¹⁶ Most of the studies on domestic VAW that have been conducted in Ethiopia, like ours, have adapted tools from existing literature, including the WHO’s domestic VAW assessment tool.^13,16–34

Cronbach’s alpha (α, or coefficient alpha) is a measure of internal consistency, or reliability, that is, how closely a set of items are related as a group. Cronbach’s alpha is developed by Lee Cronbach in 1951, which measures reliability of the tool. Reliability is how well a test measures what it should.^17,18 A review of all the studies showed that the level of Cronbach’s alpha of the domestic VAW assessment tools is not reported in most of the studies, which have been conducted in low-income countries (particularly Ethiopia). The consistency of the items of domestic VAW assessment tool is a core component of the studies, and Cronbach’s alpha is not estimated and reported. We could not find a literature that reported the reliability (Cronbach’s alpha) estimates of the domestic VAW assessment tool. Hence, the main objective of this study was to estimate the inter-item-correlation (reliability) of the tool adapted from any literature on domestic VAW in low-income country settings. This study tool was adapted from literature to assess the level of domestic VAW in the Northwestern Ethiopia. Therefore, this study may contribute to filling the literature gap of reliability estimates of tools that often used to assess domestic VAW.

Methods

Study design and setting

A community-based cross-sectional study was conducted in the Awi zone of Northwestern Ethiopia from November 15, 2017 to December 31, 2017. This was to serve as a baseline survey for a three-arm quasi-experimental study. Awi zone has nine districts, of which three districts were included in the study. It is located 447 km from Addis Ababa. According to the Awi zonal health department report published in June 2018, this zone has a total population of 1,285,242, of whom 631,054 (49.1%) are men and 654,188 (50.9%) are women. About 12.5% of the population in Awi zone live in urban areas. Almost 93.5% of the population are Ethiopian Orthodox Christian while 5.4% of the population are Muslim.¹¹ Very little is known about domestic VAW in Awi zone, but one study shows the level of VAW to be as high as 78.0%.¹³

Sample size determination and sampling procedures

Sample size was calculated using a statistical formula¹⁹ with 5% margin of error, 95% significance level, 80% power, desired intervention effect of 13%, and design effect of 1.11.¹² Eventually, the final sample size was 1,269 married or cohabitating women (15–49 years). Married or cohabitating women (15–49 years) who had lived at least 12 months with their current husband and lived at least 6 months in the selected sub-districts were eligible. Three out of nine districts were selected randomly in the Awi zone by a lottery method. Then two (urban and rural) sub-districts were selected purposefully considering their appropriateness, resource, time, and geographical non-proximity to reduce threats to validity arising from possible information contamination. Sampling frame was constructed from the health extension workers’ household registry (family-folder) to recruit eligible women from each selected sub-district. Multistage, stratified, and systematic sampling methods were used to recruit women at their permanent places of residence. The first household (random start) was recruited by lottery method using the first eligible household numbers (1 to kth value =2). In the case of two eligible women being present in a single household, one woman was selected for the interview using the lottery method (Figure 1). For further details, the protocol has been registered (ClinicalTrials.gov ID: NCT03265626) and published elsewhere.²⁰

Figure 1 Illustration of participant recruitment process.

The protocol was reviewed and approved by the Institutional Health Research Ethical Review Committee, College of Health and Medical Sciences, Haramaya University (Ref. No IHRERC/146/2017). This study was conducted in accordance with the Declaration of Helsinki,²¹ and written informed consent was obtained from each study participant (woman), and the information was kept confidential and anonymous. Confidentiality of the information was maintained, among others by avoiding personal identifiers, locking the metallic cabinet for hardcopy questionnaire and investigators placing password on computers with stored data. Participant’s deidentified data that support the analysis finding of this study as well as further analysis works will be shared as per official and valid request to the corresponding author (AS). Participant deidentified data will also be available online in the protocol registration database (ClinicalTrials.gov ID: NCT03265626), and also this journal web-pages as necessary as soon as further analysis for additional manuscripts is completed on SPSS (23.0) software after May 2019. In addition, ethical approval letter is available at any time.

Tool development and data collection methods

The data collection tool was adapted from several source in the literature^13,16–34 (Table 1). Face-to-face interviewer-administered method was carried out using the structured questionnaire administered by the 12 trained female data collectors. Data collectors’ training, pretest, and supportive supervision were provided by the principal investigator to assure the quality of data collected. Qualified female professionals (midwives, nurses, or public health workers) who have experience in field surveys and were neither resident nor deployed at nearby health facilities were hired as data collectors in order to increase the trustworthiness of the information. The training of data collectors was focused on the questionnaires, interview techniques, sampling methods, protection of confidentiality, ethical issues of domestic VAW research, and data quality assurance. Necessary amendments were made based on feedback from study participants and comments from data collectors.

Table 1 Adapted tool to measure domestic violence against women in a low-income country setting

Data processing and analysis

Overall, the domestic VAW assessment tool comprised nine sections. Three of the sections were sociodemographic and economic characteristics of women; access to sources of information about VAW and gender equality; and their husbands’ sociodemographic characteristics. The latter was not included in the reliability analysis. Six of the sections that covered the women’s decision-making potential and women’s access to household resources and control over autonomy (13 items);³⁵ women’s accepting attitude toward justified wife-beating (five items),^28,36 physical violence (seven items), psychological violence (15 items), sexual violence (five items),^37,38 and gender inequitable norm (seven items).²³ The reliability analysis was carried out for the six sections of the tool. The gathered data were entered into EpiData 3.1.0 and exported to SPSS 23.0 for further analysis. The frequency, percentage, mean, and standard deviations were computed for the participants’ sociodemographic characteristics. To examine the reliability of the tool to assess domestic VAW, the following analyses were performed: mean, standard deviation, scale mean if item deleted, scale variance if item deleted, corrected item total correlation, and Cronbach’s alpha if item deleted.

Results

The overall response rate of the survey was 95.9% (1,217/1,269). The reasons for non-response were described in detail in Figure 1. The mean age of the women was 30.0 (±7.1) years. The majority of women (98.8%, n=1,202) were formally married. Slightly more than half (52.5%, n=639) of the women were rural residents. The mean of women’s marital duration was 11.5 (±7.9) years. Furthermore, the mean age of their husbands was 37.3 (±9.3) years. About one-quarter (24.9%, n=303) were unable to read and write. About half (50.7%, n=617) engaged in trade or income-generating activities. Three-fourth (75.0%, n=913) of the women’s husbands had a history of addictive substance misuse. Of these, 99.9% (n=912) of husbands had a history of alcohol consumption. Almost one-quarter (26.4%, n=321) of the women knew their husbands’ earning (Table 2).

Table 2 Sociodemographic characteristics of women, Northwestern Ethiopia, December, 2017 (n=1,217) Abbreviation: NGO, nongovernmental organization.

Cronbach’s alpha estimate of the domestic VAW questions

Cronbach’s alpha is a measure of internal consistency (reliability) of the items in the tool, usually a scale. It shows how closely a set of items are rated as a group. It is expressed as a number between 0 and 1, the closer it is 1, the higher the reliability. Internal consistency describes the extent to which all the items in a tool measure the same concept, and hence, it is connected to the inter-relatedness of the items within the tool.⁴⁴ The overall Cronbach’s alpha of the tool was higher than the minimum recommended value of 0.70. The women’s decision-making and household resource control autonomy were assessed using 13 items and its mean was 30.2 (±5.4). The women’s accepting attitude of justified wife-beating was assessed using five items with a mean of 10.6 (±2.6). The women’s attitude toward inequitable gender-norms was assessed using seven items with a mean of 11.1 (±2.5). Cronbach’s alpha for the women’s decision-making autonomy, women’s accepting attitude toward justified wife-beating, and inequitable gender-norm were 0.764, 0.894, and 0.867, respectively. In addition, physical domestic VAW was assessed using seven items, and the mean was 12.9 (±1.8). Psychological domestic VAW was assessed using a tool with 15 items with the mean of the scale analysis of items being 27.1 (±4.0). Sexual domestic VAW was assessed using a five-item questionnaire with the mean of the scale analysis of items being 9.2 (±1.3). The Cronbach’s alphas for physical, psychological, and sexual domestic VAW assessment questions were 0.876, 0.925, and 0.812, respectively. The overall Cronbach’s alpha of the domestic VAW assessment tool was 0.785 (Table 3).

Table 3 Item characteristics, item-total correlation, and alpha if item-deleted of the different types of domestic violence against women (VAW) assessment items (n=1,217)

Discussion

This reliability analysis estimated the consistency of response from the adapted structured questionnaire(s) that were used to assess domestic VAW. Generally, the adapted survey tool had Cronbach’s alpha score of 0.785, higher than the recommended minimum of 0.70. Specifically, Cronbach’s alphas were women’s decision-making autonomy (13 items, 0.764); women’s accepting attitude of justified wife-beating (five items with 0.894); physical violence (seven items, 0.876); psychological violence (15 items, 0.925); sexual violence (five items, 0.812); and gender inequitable norm (seven items, 0.867). This tool had a Cronbach’s alpha consistent with other studies with a range of 0.68–0.80,³⁹ higher than 0.80,⁴⁰ and greater than 0.90.⁴¹ Furthermore, this finding is similar to that of a study conducted in Sweden which showed that the Cronbach’s alpha of the VAW assessment tool was higher than the minimum recommended value (>0.70).⁴²

In addition, the Cronbach’s alpha of the tool is also consistent with the tools used to assess the risk of domestic VAW in China which indicated a Cronbach’s alpha of 0.76.^43,44 However, this study finding is a bit lower than a study conducted in the USA to assess VAW which showed that Cronbach’s alpha of 0.96.⁴⁵ Nevertheless, this finding shows a relatively higher reliability than a study conducted on measurement tool used for physician assessment which has a Cronbach’s alpha of >0.65.⁴⁶ There are some arguments behind the value of Cronbach’s alpha. It is argued that it is a coefficient of the reliability or internal consistency of the items, but not a statistical test.⁴⁴ In addition, a high value for alpha does not imply that the measure is unidimensional.

The study’s finding can motivate researchers to adopt this consistent tool, which would have a great implication on the analysis of data to inform evidence-based decision-making. This is important since concrete evidence on the level of domestic VAW to understand the problem is needed to help make appropriate decisions. Therefore, this tool can be used by researchers, policy makers, clinicians, and other key stakeholders in sub-Saharan Africa and other low-income settings to enhance studies on domestic VAW. It can also be used for need assessments, program implementation monitoring, and impact evaluations.

Strengths and limitations

This study has notable strengths including it being community-based, urban–rural mix of sample, well-defined study participants, and representative sample size that can allow for generalization of findings to the general community. However, this study also has some limitations. The disclosure of domestic VAW issues can be a sensitive private issue kept as family secret in most instances. This may be affected by social desirability bias. In addition, some women may suffer from recall bias, unable to remember some of the domestic VAW experiences that they may have accepted as a part of marital life. So social desirability and recall biases may result in underreporting of domestic VAW by the study participants.

Conclusion

The adapted tool used to assess domestic VAW in Ethiopia had high reliability. Therefore, the researcher can adapt the tool for future studies. Furthermore, assessment of the reliability of the tool in other settings is recommended to confirm its applicability as a tool for low-income countries to determine the level of domestic VAW.

Availability of data and materials

The data that support the findings are available upon submitting a reasonable request to the corresponding author.

Acknowledgment

We thank Tropical Disease Research/WHO, University of Ghana, and study participants.

Author contributions

AS, KT, AM, and AA conceived and designed the study. AS carried out activities from inception to the draft of the manuscript. AS, KT, AM, NA, and AA extensively reviewed the manuscript. All authors contributed to data analysis, drafting and revising the article, gave final approval of the version to be published, and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.

---

References

1.		WHO [webpage on the Internet]. Addressing violence against women and achieving the Millennium Development Goals. Geneva, Switzerland; 2005. Available from: https://www.who.int/gender-equity-rights/knowledge/who_fch_gwh_05_1/en/. Accessed March 1, 2016.
2.		PRB. IGWG. Gender-Based Violence and Reproductive Health & HIV/AIDS. Summary of technical update. Washington, DC 20009–5728;2002. Available from: www.prb.org. Accessed March 1, 2016.
3.		USAID, PRB IGWG. Gender-Based Violence: Impediment to Reproductive Health. 2010. Available from: https://www.popline.org/node/211911. Accessed June 1, 2017.
4.		Oxfam. Ending Violence Against Women; An Oxfam Guide. 2012. Available from: https://www.oxfam.org/sites/www.oxfam.org/files/ending-violence-against-women-oxfam-guide-nov2012.pdf. Accessed November 1, 2015.
5.		WHO. Violence Against Women and HIV/AIDS: Critical Intersections Intimate Partner Violence and HIV/AIDS. WHO Bulletin Series. 2004;(1):1–9.
6.		The World Student Christian Federation and the world YWCA. Domestic Violence Fact Sheet. A 2010 Lenten Study Compiled by the World Council of Churches; 2010. Available from: http://women. overcomingviolence.org. Accessed May 1, 2017.
7.		WHO. Global and regional estimates of violence against women: prevalence and health effects of intimate partner violence and non-partner sexual violence. Geneva, Switzerland; 2013. Available from: www.who.int/reproductivehealth. Accessed December 1, 2015.
8.		Boldosser-Boesch A, Byrnes D, Carr C, et al. Sexual and Reproductive Health and Rights (SRHR) and the Post- Development Agenda. Brief Cards. 2015. Available from: https://www.scribd.com/document/235297283/UAP-SRHR-Post-2015-Briefing-Cards. Accessed June 1, 2016.
9.		Ellsberg M, Jansen HAFM, Heise L, Watts CH, García-Moreno C, Study W. Intimate partner violence and women’s physical and mental health in the WHO multi-country study on women’s health and domestic violence: an observational study. Lancet. 2008;371(9619):1165–1172.
10.		Semahegn A, Mengistie B. Domestic violence against women and associated factors in Ethiopia; systematic review. Reprod Health. 2015;12(1):78.
11.		CSA, Central Statistics Agency. Ethiopia Demographic and Health Survey (2011): Central Statistical Agency Addis Ababa, Ethiopia. ICF International Calverton, Maryland, USA; 2012. Available from: https://dhsprogram.com/pubs/pdf/SR191/SR191.pdf. Accessed October 1, 2015.
12.		CSA, Central Statistical agency. Ethiopia Demographic and Health Survey 2016. Addis Ababa, Ethiopia, and Rockville, Maryland, USA: CSA and ICF; 2016. Availbale from: www.ethiodemographyandhealth.org/Measure_DHS_Ethiopia2016.pdf. Accessed June 1, 2017.
13.		FDRE, Constitution of the Federal Democratic Republic of Ethiopia Constitution of The Federal Democratic Republic of Ethiopia. Addis Ababa, Ethiopia; 1994. Available from: www.wipo.int/edocs/lexdocs/laws/en/et/et007en.pdf. Accessed June 1, 2017.
14.		FDRE, The Criminal Code of the Federal Democratic Republic of Ethiopia 2004. Proclamation No 414/2004. Addis Ababa, Ethiopia: Federal Democratic Republic of Ethiopia; 2005. Available from: www.wipo.int/edocs/lexdocs/laws/en/et/et011en.pdf. Accessed February 1, 2018.
15.		FDRE. The Revised Family Code Proclamation No 213/2000. Addis Ababa, Ethiopia; 2000:96. Available from: hrlibrary.umn.edu/research/family code (English).pdf. Accessed September 1, 2017.
16.		FMOH. Standard Operation procedure for the response and prevention of sexual violence in Ethiopia: Federal Democratic Republic of Ethiopia Ministry of Health. Addis Ababa, Ethiopia; 2016.
17.		Tavakol M, Dennick R. Making sense of Cronbach’s alpha. Int J Med Educ. 2011;2:53–55.
18.		Idre [webpage on the Internet]. What does Cronbach’s alpha mean? Institute for Digital Research and Education, UCLA; 2018. Available from: https://stats.idre.ucla.edu/spss/faq/what-does-cronbachs-alpha-mean/. Accessed June 1, 2018.
19.		Biresaw A. Awi zone health department annual report of the population with in the zonal administration catchments on June 30th, 2018. Zonal reproductive health team leader (received via phone call). Injibara, Ethiopia; 2018.
20.		Semahegn A, Belachew T, Abdulahi M. Domestic violence and its predictors among married women in reproductive age in Fagitalekoma Woreda, Awi zone, Amhara regional state, North Western Ethiopia. Reprod Health. 2013;10(1):63.
21.		Zhong B. How to Calculate Sample Size in Randomized Controlled Trial? J Thorac Dis 2009. 2009;1(1):51–54.
22.		Semahegn A, Torpey K, Manu A, Assefa N, Ankomah A. Community based intervention to prevent domestic violence against women in the reproductive age in northwestern Ethiopia: a protocol for quasi-experimental study. Reprod Health. 2017;14(1):155.
23.		World Medical Association (WMA). World Medical association Declaration of Helsinki ethical principles for medical research involving human subjects, 5th (Edinburgh) Amendment and note of clarification. JAMA. 2013;310(20):2191–2194.
24.		Nguyen PH, Nguyen SV, Nguyen MQ, et al. The association and a potential pathway between gender-based violence and induced abortion in Thai Nguyen Province, Vietnam. Glob Health Action. 2012;5(19006):11.
25.		Abebe Abate B, Admassu Wossen B, Tilahun Degfie T, Abate BA, Wossen BA, Degfie TT. Determinants of intimate partner violence during pregnancy among married women in Abay Chomen district, Western Ethiopia: a community based cross sectional study. BMC Womens Health. 2016;16(1):16.
26.		Rahman M, Nakamura K, Seino K, Kizuki M. Does gender inequity increase the risk of intimate partner violence among women? Evidence from a national Bangladeshi sample. PLoS One. 2013;8(12):e82423.
27.		Sapkota D, Bhattarai S, Baral D, Pokharel PK. Domestic violence and its associated factors among married women of a village Development Committee of rural Nepal. BMC Res Notes. 2016;9(1):178.
28.		Lamichhane P, Puri M, Tamang J, Dulal B. Women’s status and violence against young married women in rural Nepal. BMC Womens Health. 2011;11(1):19.
29.		Dalal K, Wang S, Svanström L. Intimate partner violence against women in Nepal: an analysis through individual, Empowerment, family and societal level factors. J Res Heal Sci J. 2014;14(4):251–257.
30.		Uthman OA, Moradi T, Lawoko S. Are individual and community acceptance and witnessing of intimate partner violence related to its occurrence? Multilevel structural equation model. PLoS One. 2011;6(12):e27738.
31.		Shannon K, Leiter K, Phaladze N, et al. Gender inequity norms are associated with increased male-perpetrated rape and sexual risks for HIV infection in Botswana and Swaziland. PLoS One. 2012;7(1):e28739.
32.		Abramsky T, Watts CH, Garcia-Moreno C, et al. What factors are associated with recent intimate partner violence? Findings from the WHO multi-country study on women’s health and domestic violence. BMC Public Health. 2011;11(1):109.
33.		Ismayilova L. Spousal violence in 5 transitional countries: a population-based multilevel analysis of individual and contextual factors. Am J Public Health. 2015;105(11):e12–e22.
34.		Hayati EN, Högberg U, Hakimi M, Ellsberg MC, Emmelin M. Behind the silence of harmony: risk factors for physical and sexual violence among women in rural Indonesia. BMC Womens Health. 2011;11(1):52.
35.		Antai D. Traumatic physical health consequences of intimate partner violence against women: what is the role of community-level factors? BMC Womens Health. 2011;11(1):56.
36.		Uthman OA, Lawoko S, Moradi T. Factors associated with attitudes towards intimate partner violence against women: a comparative analysis of 17 sub-Saharan countries. BMC Int Health Hum Rights. 2009;9(1):14.
37.		Conroy AA. Gender, power, and intimate partner violence: a study on couples from rural Malawi. J Interpers Violence. 2014;29(5):866–888.
38.		Heise LL. Violence against women: an integrated, ecological framework. Violence Against Women. 1998;4(3):262–290.
39.		Ellsberg M, Heise L. Researching Violence against Women; A Practical Guide for Researchers and Activists. Washington DC, United States: World Health Organization and PATH; 2005:259.
40.		Emami A, Safipour J. Constructing a questionnaire for assessment of awareness and acceptance of diversity in healthcare institutions. BMC Health Serv Res. 2013;13(1):145.
41.		Jenkinson C, Coulter A, Wright L. Short form 36 (SF36) health survey questionnaire: normative data for adults of working age. BMJ. 1993;306(6890):1437–1440.
42.		Shakil A, Donald S, Sinacore JM, Krepcho M. Validation of the hits domestic violence screening tool with males. Fam Med. 2005;37(3):193–198.
43.		Svalin K, Mellgren C, Levander MT, Levander S. The inter-rater reliability of violence risk assessment tools used by police employees in Swedish police settings. Nord Polit. 2017;4(1):9–28.
44.		Chan KL. Predicting the risk of intimate partner violence: the Chinese risk assessment tool for victims. J Fam Violence. 2012;27(2):157–164.
45.		Chan KL. Assessing the risk of intimate partner violence in the Chinese Population: The Chinese Risk Assessment Tool for Perpetrator Assessing the Risk of Intimate Partner Violence in the Chinese Population: The Chinese Risk Assessment Tool for Perpetrator (CRAT). Violence Against Women. 2014;20(5):500–516.
46.		Dienemann J, Glass N, Hanson G, Lunsford K. The domestic violence survivor assessment (DVSA): a tool for individual counseling with women experiencing intimate partner violence. Issues Ment Health Nurs. 2007;28(8):913–925.