WOMEN’S ATTITUDE TOWARDS VIOLENCE AGAINST WOMEN IN RELATION TO LEGAL AID SERVICE INTERVENTIONS IN MOROGORO RURAL AND KONGWA DISTRICTS, TANZANIA

Tatu M. Nyange¹⁺----Anna N. Sikira²----Joyce G. Lyimo-Macha³

¹Department of Gender Studies, the Mwalimu Nyerere Memorial Academy, Dar es Salaam, Tanzania

²Department of Development Studies, Sokoine University of Agriculture, Morogoro, Tanzania

³Institute of Continuing Education, Sokoine University of Agriculture, Morogoro, Tanzania

Keywords:GBV against women Legal aid services (LAS) Women’s attitude Tanzania

ARTICLE HISTORY: Received:24 March 2017, Revised:29 May 2017, Accepted:14 June 2017 Published:: 24 July 2017.

Contribution/ Originality:The paper's primary contribution is finding that ethnic background and women involvement in Legal Aid Service (LAS) interventions; are found to be stronger predictor and important determinant factors on influencing attitude towards Gender Based Violence against women.

1. INTRODUCTION

Gender Based Violence (GBV) against women is described as the most prevalent human rights violation in the world (Arango et al., 2014). Various theories explain the origin of GBV with different perspectives, for example socio-cultural theories of violence believed that GBV is a result of reflection of attitude shared by a group of people governing interpersonal interactions in patriarchy societies (Nayaki et al., 2003; Anderson, 2005). While, literatures indicate that women are more affected by Gender Based Violence (GBV) in their intimate relationship compared to men (FEMNET, 2001; Arango et al., 2014) Yet, a study conducted by Sikira et al. (2010) in Serengeti District in Tanzania reported that the majority of women had positive attitude towards GBV against them, which is likely to slow down the process of eradicating GBV. In many societies, prevailing attitudes subordinate women to men and entitle men to use violence to control women. These attitudes serve to justify, tolerate or condone violence against women (World Health Organization (WHO) and Pan Africa Health Organization (PAHO), 2012). It is therefore this current study focuses on attitude towards GBV against women.

Violence against Women (VAW) as defined by United Nation (UN) (1993) is "any act of GBV that results in, or is likely to result in, physical, sexual or mental harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life." There are different forms of violence against women, either in the form of physical, sexual, psychological and economic (Nair, 2014).

Literature reveals that the most common forms of GBV against women practiced in Tanzania take different forms throughout women's and children's lives, including: wife beating, sexual violence such as marital rape, deprivation of basic necessities, early marriage, abuse of elderly, cultural practices like FGM, cleansing of widows/widowers and marriage between two women “nyumba nthobhu1” (Sikira et al., 2010; NBS and ICF Macro, 2011; LHRC, 2012; LHRC, 2013) also Nyange et al. (2016) reported that, wife beating, deprivation of basic necessities and marital rape are the most common forms of GBV against women practiced in Kongwa and Morogoro Rural districts, Tanzania.

For the initiatives to eliminate GBV against women to be successful, it is crucial to change women's attitude, because positive attitude towards GBV is associated with actual occurrence of violence (Antai and Antai, 2008).

The Government of Tanzania has made efforts in tackling GBV against women through ratification of legal instruments on the rights of women. These include:  the Convention for the Elimination of All forms of Discrimination Against Women (CEDAW),  the Sexual Offences Special Provision Act of 1998 (SOSPA) aiming at punishing the perpetrators of violence and the Protocol to the African Charter on Human and Peoples’ Rights on the Rights of Women in Africa (Maputo Protocol). Also various Legal Aid Services (LAS) 2 providers were established with the aim of enhancing awareness of women on human legal rights including all forms of GBV against women such as FGM (Legal and Human Right Centre (LHRC), 2004) .

This study assumed that intervention programmes undertaken by LAS providers and other development actors, through dissemination of information about women legal rights and harmful practices such as FGM would enhance knowledge on the effect of GBV against women among victims and perpetrators. Consequently, the knowledge contributes to one’s attitude towards GBV against women. This assumption concurred with Ajzen (1991) who assumes that humans are essentially rational and can use information available to them to make reasonable behavioral decisions; such information induces one’s attitude towards certain behaviour. In line with this background it is important to assess determinants of women attitudes towards GBV against women in context of LAS intervention.

Flood and Pease (2009) highlighted various factors influencing attitude of women towards violence against women such as age, witness of violence, religion and education campaign. Several studies have been conducted on attitude of women towards GBV against women in Tanzania. For example, a study by Abeid et al. (2015) determined knowledge of and attitude towards sexual rape and sexual abuse; while Sikira (2010) examined attitude of women towards GBV. However, all aforementioned studies did not focused on attitude of women towards GBV against women in the context of LAS interventions. The phenomenon may be necessarily something new towards elimination of GBV against women. This study therefore aimed at examiningwomen's attitude towards GBV against women in relation to their involvement in LAS interventions in Kongwa and Morogoro Rural districts. Specifically, the study assessed attitude of women towards GBV against women among beneficiaries and non-beneficiaries of LAS interventions, analysed determinants factors influence attitude of women towards GBV against women in relation to LAS interventions and tested the hypothesis that: Attitudes towards GBV against women is significantly different between beneficiaries and non-beneficiaries of LAS interventions.

Changing victims and perpetrators’ attitude towards GBV is highlighted as a stepping-stone towards elimination of GBV (Sikira et al., 2010). Likewise, Nayaki et al. (2003) informed that an understanding of attitude toward GBV against women is crucial for effective prevention strategies.  The study aimed at contributing to the implementation of Sustainable Development Goal (SDG) Number 5 (Osborn et al., 2015) which focuses on fighting VAW as a way of achieving gender equality. Also, the study is aligned with Maputo Protocol Article 8 (a) and (b) which promotes women’s rights in relation to LAS (Legal Services Facility (LSF), 2012). It is anticipated that the findings of this study will provide insights to policy makers on the need to plan interventions for creating awareness on women's legal rights, which can change attitudes of women towards GBV against women. The findings enlighten gender activists, LAS providers and the government designing and execute LAS towards eliminating GBV against women.

2. METHODOLOGY

2.1. Study Areas

The study was conducted in Morogoro Rural District located in Morogoro Region and Kongwa District located in Dodoma Region. The two regions are within top five (Dodoma, Mara, Ruvuma, Morogoro, and Kagera) regions with high prevalence of physical violence in 2010 (NBS and ICF Macro, 2011).

Morogoro Rural and Kongwa districts were purposively selected due to presence of organisations/networks actively running LAS interventions at the time of the study. Some of the networks include: Anti- Female Genital Mutilation Network (AFNET), Tanzania Women Lawyers Association (TAWLA), Morogoro Paralegal Centre (MPC) and a group of legal aid services providers in Mvuha “Kikundi cha Wasaidizi wa Kisheria Mvuha” (KIWAKIM) which advocates on human and women's legal rights including issues related to GBV against women. Demographically, women comprise 51.9% out of the 309 973 people and 50.8% out of the 286 248 people in Morogoro Rural and Kongwa districts respectively (United Republic of Tanzania (URT), 2013). The main ethnic groups in Kongwa are Wakaguru and Wagogo, while Wakutu, Waluguru and a small proportion of Maasai and Wasukuma migrants are found in Morogoro Rural District.

Likewise, the selected study areas differ in terms of their cultural backgrounds; Kongwa District mainly practices a patrilineal system while Morogoro Rural District mainly practices a matrilineal system. According to these characteristics, the districts are expected to be essential in studying attitude of women towards GBV against women, as perceptions from the different cultural backgrounds can be enlightening.

2.2. Research Design and Sampling Technique

In order to collect multiple cases in a single point of time the study employed a cross-sectional research design (Bailey, 1998). The design is considered to be appropriate in providing an image of a current situation in a specific time (Ellsberge and Heise, 2005). One division, one ward and four villages were purposely selected from each district based on the presence of organisations or networks actively running LAS interventions; also in the same division one ward and four villages were purposely selected from each District based with absence of LAS intervention.

The sampling unit for this study was an individual woman involved and those not involved in LAS intervention activities. Only women were involved in this study because women are the most affected group by different forms of violence and are the one’s actively involved in LAS interventions activities. For the purpose of the current chapter, a woman involved in LAS interventions is regarded as a beneficiary of LAS interventions. This means that they are affiliated in an organization/group of LAS providers at village level or should have attended educational training on women legal rights and GBV issues organized by LAS providers. Women who are not involved in LAS intervention are regarded as non-beneficiaries.

In addition, women who accessed LAS interventions such as those who were counseled on any issues such as matrimonial cases, and land conflict offered by LAS providers are not regarded as beneficiaries of LAS interventions as they lack aspect of attending educational training on women rights and other related sessions. The population for this study was women aged from 15 years and above with different marital statuses (single, married, cohabitating, separated, divorced and widowed).

Eight villages were purposively selected where four villages were composed of women who were beneficiaries of LASs, while four villages were composed of women who were non-beneficiaries of LAS interventions. From each village, 30 women were selected. Selection of villages also based with the following consideration such as cultural context, spillover effect and village within the same district shared a common system such as formal or informal justice systems and others alike. Sampling frame for LAS beneficiaries was obtained from a list of names registered in LAS intervention networks or groups with the assistance of LAS facilitators at village level. While for non-beneficiaries the sampling frame was a list of names of all women in non-intervention villages obtained with assistance of village leaders. Simple random sampling using the lottery method was used to select sample from the list of names selected from each village. The sample size was of 240 respondents, whereby 120 respondents were beneficiaries of LAS interventions and 120 were non-beneficiaries.

2.3. Data Collection

Both primary and secondary data were collected. Key Informant Interviews (KII) and Focus Group Discussions (FGDs) were used as the main source of qualitative data. An interview guide was used to gather information from 24 key informants (four representatives from LAS providers, two police officers from nearest police stations working at the gender desk, two District Community Development Officers, four members from ward tribunals, four Ward Officers and eight Village Executive Officers. A focus group interview guide was used in discussion to gather information from 16 FGDs which involved 12 people in each discussion session (two FGDs of women and men separately from each village). This was because the study learned from one FGD in Mtamba village (During pre-testing of instrument) where men and women were combined in one group whereby women were not able to speak freely in the presence of men.

FGD, KII and in-depth discussions were used to provide additional and detailed information to explore attitude towards GBV against women in the study areas. While a structured questionnaire with close-ended questions was used to collect quantitative data about attitude towards GBV against women. Pre-testing of the instrument for reliability was done in Mtamba village near Mvuha village in Morogoro Rural District which involved twenty respondents.

Attitude of women towards GBV against women was measured by using a Likert scale. The scale has been found to be an effective technique for the measurement of attitudes (Likert, 1932). The study used a 5 points Likert scale (1 = strongly agree, 2 = agree, 3 = undecided, 4 = disagree and 5 = strongly disagree). Responses from all statements were combined to create a measurement of an Attitudinal Scale (AS). AS is a single variable used to represent cumulative perception of attitude (Likert, 1932).

Statements favorable to the construct were positively worded while unfavorable statements used negative connotation. Then numerical values for the response options were reversed when calculating the overall score. The higher values indicated positive attitude disposition of favorably attitude towards GBV against women, implying that the respondent was supporting violence practices against women. While low values indicated negative attitude (i.e. unfavorably response) towards GBV against women, implying that the respondents were not supporting violence practices against women.

Field (2009) suggests that, it is useful to check the reliability of a scale and the most common measure of scale reliability is Cronbach’s alpha. Reliability gives indication of an instrument whether it can be interpreted consistently across different situations. Cronbach’s alpha value ranges from 0 to 1.0 (De Vaus, 2002). Scholers suggests that a value of 0.7 to 0.8 is an acceptable value for Cronbach’s alpha (Pallant, 2007; Field, 2009). The Cronbach’s alpha for attitude scale in this study was 0.74, which indicates that the research instrument was reliable.

The overall scores on the Likert scale were categorized into positive, neutral and negative attitude towards GBV against women. The highest possible score was calculated by multiplying 8 statements by 5 points to get 40 points; while the middle point was calculated by multiplying 8 statements with 3 points to get 24 points, and the lowest possible score was calculated by multiplying 8 statements by 1 point to get 8 points. Therefore, 24 was the cut-off point and stood for neutral attitude. Hence, scores from 8 to 23 on the overall scores were considered as negative attitude; while 25 to 40 stood for positive attitude.

A Mann Whitney U test was used to test the hypothesis that attitude towards GBV against women for women beneficiaries and non-beneficiaries of LAS intervention were significantly different (p≤0.05). The test was ideal because the technique is appropriate to test for differences between two independent groups (Pallant, 2007). Also the test is the alternative of the t-test for independent samples and also it analyses differences in the positions of ordinal dependent variables in two independent groups (Nachar, 2008).

The ordinal logistic regression was used to determine LASs related factors on attitude of women towards GBV against women. The reason for using ordinal regression model was due to the fact that the dependent variable measured at the ordinal level in terms of ranked alternative responses (Negative attitude, neutral attitude and positive attitude). The independent variables included socio-demographic variables as indicated in the ordinal logistic regression model below. Interpretation of the output from the model focused on p-values (P≤0.05) considered be statistically significant for testing the significance of the effect; Wald statistics for measuring the strength of the effect and coefficients for measuring the directions of the effect. The ordinal logistic regression model used in this study is presented in Equation (i). The model was adopted from Agresti and Finlay (2009) was:

Where:

P(Y) = the probability of the success alternative occurring, e = the natural log, α = the intercept of the equation, β1 to βk = coefficients of the predictor variables, X1 to Xk = predictor variables entered in the ordinal regression model. Specifically in this study: P(Y) = the probability of women being grouped in the negative attitude towards GBV against women; α = the intercept of the equation; b1… b10 = Regression coefficients; X1… Xk predictor or independent variables entered in the model, which were: X1 = age of respondent (measured in years), X2 = education level of respondent measured in years of schooling, X3 = experience of violence at adulthood (Victim of violence at adulthood 1, 0 otherwise),  X4 = witness of violence during childhood (witnessed violence during childhood 1, 0 otherwise), X5 = ethnic background (matrilineal 1 and patrilineal 0), X6 = access of LAS ( received LAS 1 and 0 otherwise), X7 = experience of violence at childhood (Victim of violence at childhood 1, 0 otherwise),  X8 = involvement into LAS interventions activities (Beneficiary of LAS 1, non-beneficiary of LAS 0),  X9 = awareness of women rights (total scores on women legal right index), X10 = religion (being affiliated in Christian or Muslim faith 1, 0 otherwise).

3. RESULTS AND DISCUSSION

3.1. Profile of the Respondents

The findings presented in Table 1 show that almost one third (29.6%) of women were in the age group between 25 and 34 years of age including 15% and 14.6% women from Morogoro Rural and Kongwa districts respectively. The study findings show that 45.4% of the women completed primary education while 21.2% have no formal education, and 20% did not complete their primary education (Table 1). Since education is regarded as one of the major components in shaping one’s attitude, the trend in the study areas can cause difficulties in extracting information from brochures and other reading materials, which are mostly used by LAS providers to disseminate information about women legal rights and other issues related to GBV against women. Therefore, lack of or low education might constitute one’s attitude towards wrong perception.

With regard to ethnic background, the findings show that more than one-third (36.7%) of the women were Wakaguru and 5.8% were Wagogo, mainly from Kongwa District. The two ethnic groups essentially practise a patrilineal system. On the other hand 18.8% were Wakutu and 14.6% were Waluguru from Morogoro Rural District who follow a matrilineal system. While 15.3% of respondents were other ethnic groups including: Wahehe, Wasukuma, Wazaramo and Wazigua. Small proportions (8.8%) of the respondents were the Masaai who migrated from Arusha and Manyara regions to Morogoro Rural District. The findings indicate a significant variation of ethnic groups with different cultural background, implying that capturing perception of different forms of VAW from diverse cultural backgrounds is likely to be enlightening, which is essential in studying attitude towards GBV against women.

Table-1. Distribution of respondents by age, education and ethnic profile (n = 240)

Variable	Kongwa		M R		All
	n	%	n	%	n	%
Age
15-24	28	11.7	11	4.6	39	16.2
25-34	35	14.6	36	15.0	71	29.6
35-44	35	14.6	27	11.2	62	25.8
45 +	22	9.2	46	19.2	68	28.3
Education level
Completed Primary education	63	26.2	46	19.2	109	45.4
Not attended school	14	5.8	37	15.4	51	21.2
Not completed primary education	25	10.4	23	9.6	48	20.0
Completed secondary education	13	5.4	10	4.2	23	9.6
Not completed secondary education	4	1.7	3	1.2	7	2.9
Completed technical or diploma education	1	0.4	1	0.4	2	0.8
Ethnic groups
Wakaguru	88	36.7	0	0.0	88	36.7
Wakutu	0	0.0	45	18.8	45	18.8
Waluguru	1	0.4	34	14.2	35	14.6
Wamaasai	0	0.0	21	8.8	21	8.8
Wagogo	14	5.8	0	0.0	14	5.8
Others ( Wazigua, Wahehe, Wazaramo, etc)	17	7.1	20	8.4	37	15.3

Note: MR = Morogoro Rural, All = Kongwa and Morogoro Rural districts

3.2. Attitude towards GBV against Women among Beneficiaries and Non-Beneficiaries of LAS

The findings presented in Table 2 indicate common myths which are believed to support and others are in opposition to GBV against women. Respondents were required to show their attitude towards GBV against women by indicating their degree of agreement with a set of statements which support GBV practices against women. The majority (82.9%) of the women (of these  45.4% were beneficiries and 37.5% non-beneficiries of LAS intervention) disagreed with the statement that “wife battering is a sign of love”. However, one male FGD discussant from Nghumbi village in Kongwa District argued that: “…wife beating could be a sign of love and not a sign of hatred... ” (Male FGDs Nghumbi Village).

He pinpointed that if a wife misbehaves she should be punished by her husband. This is because  they believe that men are the protectors and controllers of women, as men were traditionally given the mandate to control women. This implies that perpetrators maintaining violent practices through patriachal system that gives men power to control women under justification of women protection. Likewise, this is linked to the cultural based theory, which asserts that the power of tradition and norms within African culture are the source of GBV perpetuation (Merry, 2006). Since culture is a stable pattern of beliefs, traditions and values are passed from one generation to another through socialization. In African culture boys are socialized to control the family, specifically girls and women, while girls are raised to be dependent, submissive and obtain security from men (Anderson, 2005).

The findings in this paper show that majority (81.7%) of the women disagreed that it is necessary to undergo FGM in order to prepare a girl before marriage (Table 2). Moreover, the findings indicate that 57.1% of the women interviewed supported the statement that FGM is violation of women's rights and hence should be abolished. This was a good indication that women were conscious on FGM as inhumanity practice. Despite the good indication an in-depth interview with a woman in Nghumbi village denoted that FGM is important for a girl prior to marriage; she said:

“…FGM is essential for initial preparation of a girl before marriage...” (Woman in-depth interview, Nghumbi Village)

Also, she highlighted that the practice is an indication of a girl’s readiness for marriage, whereby girls get opportunities of being taught about ethnic laws, morality, good behaviours towards elders, obedience to their expected husbands and being good wives. This finding implies that, even though the majority of women in the study areas showed negative attitude towards FGM, still some of them uphold such a harmful traditional practice. This is done in order to please men as during FGM ceremonies, girls are taught to become good mothers.

Furthermore, the findings indicate that 73.8% of the women interviewed agreed that GBV against women should be reported for legal action including 45% beneficiaries and 27.9% non-beneficiaries of LAS interventions. Likewise, 65.4% of women disagreed that GBV against women is  private issue and no one should intervene (Table 2). This is in line with findings by Sikira et al. (2010) who reported that 68.4% of women in Serengeti District disagreed with false statement that GBV is a private or family issue and hence that it should be reported for legal action. This implies that women are likely to report violence incidences to appropriate legal enforcement authorities which are crucial to the success of eradication campaign of GBV against women. Although findings from one of the key informant (LAS provider) in Kongwa District provide different depiction that:

“…rigid traditions in Kongwa District are dominated by patriarchy; married women are sanctioned to disclose any type of violence to their in-laws. This is unlikely for women to report their violence issues to formal or informal legal authorities…”

The findings in Table 2 show that 66.7% of the women agree with the statement that husband/partner has no right to beat his spouse in any circumstances. The finding implies that denial of such practice reflects women's recognition of their rights and perception that no one can take action against others before the law.

Table-2. Respondents attitudes towards GBV against women (n = 240)

Statements	Agree (%)			Neutral (%)			Disagree (%)
	B	N-B	T	B	N-B	T	B	N-B	T
GBV against women is a private issue	8.8	21.7	30.4	1.7	2.5	4.2	39.6	25.8	65.4
Wife battering is a sign of  love	2.5	7.5	10.0	2.1	5.0	7.1	45.4	37.5	82.9
Women  should tolerate GBV against them in order to maintain harmony in their marital relationship	15.0	33.8	48.8	3.3	10.0	13.3	31.7	6.2	37.9
Husband/partner has no rights to beat his wife in any situation	38.3	28.3	66.7	3.8	7.5	11.3	7.9	14.2	22.1
GBV against women should be reported for legal action	45.8	27.9	73.8	1.2	10.0	11.2	2.9	12.1	15.0
Bride price promote  inequality in marriage	28.8	9.2	37.9	2.1	7.5	9.6	19.2	33.3	52.5
FGM is necessary for preparation of respectable marriage	0.8	9.6	10.4	3.8	4.2	7.9	45.4	36.2	81.7
FGM is violation of women rights should be eliminated	19.2	37.9	57.1	2.1	9.6	11.7	10.0	21.2	31.2

Note: Figures in (%) represent multiple responses, B=Beneficiary, N-B= Non-Beneficiary and T=All (B+NB)

Furthermore, the findings show that 52.5% of  the women indicated their acceptance of bride price being paid before marriage, by negating the statement that bride price promotes inequality in marriage (Table 2). This implies that woman in the study areas supported the culturally embedded act of paying bride price before marriage. Participants during FGD contended that payment of bride price is a custom relating to dowry or exchanging of gifts before or after marriage. The findings are in line with those by Sikira et al. (2010) which indicated that majority (84%) of the women in Serengeti District strongly, concurred with payment of bride price.

3.3. General Attitude of Respondent towards GBV against Women

The findings in this paper indicate that the overall attitude of women towards GBV against women was 57.1% showing negative attitude towards GBV against women including 42.5% beneficiaries and 14.6% non-beneficiaries of LAS intervention. One-third (35.4%) of the respondents have positive attitude towards GBV against women, including 33.8% non-beneficiaries and 1.7% who were beneficiaries of LAS interventions (Fig. 1). These findings indicate that majority of the non-beneficiaries of LAS uphold positive attitude towards GBV against women compared to their counterparts.

The differences in attitude among beneficiaries and non-beneficiaries of LAS intervention could be as a result of interventions done by the government, LAS providers and other women activists through dissemination of information about women’s legal rights and different forms of GBV against women. On the other hand positive attitude could be contributed by women's unawareness and lack of information on GBV against women issues.  These findings have implications to LAS providers, specifically to information offered among beneficiaries about women legal rights and harmful practices, such as FGM, which consequently constitute to their attitude towards GBV against women. The finding is aligned with the Theory of Reasoned Action (TRA) which postulates that human beings are rational and can use the available information to make reasonable behavioral decisions (Ajzen, 1991).

Source: Survey data (2013)

3.4. Differences of Respondents Attitude towards GBV against Women among Beneficiaries and Non-Beneficiaries of LAS Interventions

Mann-Whitney U test was conducted to test the difference of women’s attitude towards GBV against women between beneficiaries and non-beneficiaries of LAS interventions. The results indicate significant (p≤0.05) difference in median scores of 16 and 25 for beneficiaries and non-beneficiaries of LAS intervention respectively (Table 3). The “r” value was 0.62, which represents huge effect as the effect was above the 0.5 threshold for a large effect (Field, 2009). Thus the hypothesis that women beneficiaries and non-beneficiaries of LAS intervention are likely to differ in their attitude towards GBV against women is confirmed. Low scores of median among beneficiaries signify respondents’ negative attitude towards wrong myths that favour GBV against women practices (Table 3). This finding was also confirmed by respondents in two FGDs of women beneficiaries in Mvuha and Lukulunge villages pointed out that:

“Sensitization about women’s right and different forms of VAW offered by LAS providers was empowering them to understand false myths that support GBV against women...” (Women FGDs in Mvuha and Lukulunge Village)

Likewise, during the two FGDs of women beneficiaries of LAS interventions in Mlali-iyegu and Iduo villages, participants explained that: when they attended maternity clinics, sometimes LAS providers conducted training about women legal rights on sexual and reproductive health rights and harmful traditional practices such as FGM. Women who attended the sessions appreciated that information helped them to understand the negative impact of GBV against them and influencing their behaviour against positive attitude towards violence. However, FGD discussants in men group in Mlali-iyegu village complained that training offered by LAS providers teaches their wives to be disobedient against them. One male focus group discussant from Mlali- iyegu argued that:

“…Before my wife started to attend maternity clinic she was very obedient to me; she listened to me and followed my instructions. We lived in peace and harmony according to our traditions. We believed that men are born to control women and women should obey orders from their husbands. Since my wife started to attend those stupid campaigns offered by LAS providers in clinic, she started challenging me and ignores our traditions by saying that men and women are equal. This is nonsense and lack of respect; it does not sound in my mind….” (Male FGD Mlali-iyegu Village)

This finding implies that men have negative attitude towards training offered by LAS providers, as they were trying to maintain their patriarchal perceptions that makes men superior to women under justification of traditions. The findings have negative implications to LAS providers on changing men perceptions.

Table-3. Attitude towards GBV against women among beneficiaries and non-beneficiaries of LAS interventions (n=240)

Respondents'	n	Median	Mann-Whitney  U	Z	P-Value
Beneficiaries	120	16.00	2019.000	- 9.656	0.000
Non beneficiaries	120	25.00

r value = -0.62

3.5. Determinants of Women Attitude towards GBV against Women

The previous section 3.4 in this paper indicates that a greater proportion of women in Morogoro Rural and Kongwa districts have negative attitude towards GBV against women. Using Ordinal Logistic Model it is important to examine which variables played a great effect on altering the attitude of women towards GBV against women. The overall model fitting information showed statistically significant chi-square statistic (p<0.000). This indicates that the model gave better predictions of the outcome categories. Goodness of Fit test showed (p>0.05) which indicates data and the model predictions were similar, implying good model.

The Pseudo R-Square was 0.34 Cox and Snell and 0.409 Nagelkerke, implying that independent variables entered in the model explained 34% and 41% respectively of the variance on women attitudes towards GBV against them. The test of parallel lines showed (P> 0.05) which led to none rejection of the null hypothesis of parallelism. Most of the output from the OLM focused in this chapter were: p-values for testing the significance influence of the variables, positive or negative sign of coefficients indicate directions of women’s attitude being grouped in high category (negative attitude) or low category (positive attitude) towards GBV against women; while Wald statistics determined the strength of the influence on women attitudes towards GBV against women.

The empirical findings of the Ordinal Logistic Regression model show that three out of 10 independent variables (Involvement of women in LAS interventions, ethnical background and witness of violence at childhood) were the most significant factors on influencing women’s attitude towards GBV against them (p≤ 0.05) (Table 4).The findings show that involvement of women in LAS intervention had a significant and stronger (Wald = 26.269) influence on attitude of women towards GBV against women. The negative coefficient (Table 4) implies that women who were not involved in LAS interventions were less likely to have negative attitude than those who were involved in the interventions. This means that involvement of women in LAS interventions as empowerment strategy is essential in changing wrong myth against GBV against women.

On the other hand, results show that ethnic background had a strong (Wald=19.821) influence on attitude of women towards GBV against women. The negative coefficient of ethnic background indicates that women who belong to patrilineal ethnic background were less likely to have negative attitude relative to those with matrilineal ethnic background (Table 4). This was an expected finding in Morogoro Rural and Kongwa  districts, as the selected study areas had significant variation in ethnic groups with different cultural background (patrilineal and matrilineal), which resulted to different perception. Literally, most patrilineal societies favour GBV practices against women in order to maintain their perpetration towards GBV against women as well as domination of women (Merry, 2006).

The results further indicate positive coefficient in witness of violence at childhood; implying that those who had not witnessed violence at childhood their possibility of negative attitude towards GBV against women increases. This means women who witnessed violence incidences at their childhood become more likely to have positive attitude towards GBV against women. This finding is in line with other researchers elsewhere who reported that witnessing violence while growing up has direct impact on attitudes which in turn impacts on violence perpetration  (Mac Kowitz, 2001cited in; Flood and Pease (2009)).  

Table-4. Determinants of attitude towards GBV against women (n = 240)

	B	SE	Wald	Sig
Age	0.006	0.013	0.195	0.659
Education	-0.024	0.048	0.242	0.623
Experienced  GBV at adulthood	-0.576	0.418	1.905	0.168
Witness GBV at childhood	0.772	0.331	5.443	0.020
Ethnic  background	-1.623	0.364	19.821	0.000
Access of LAS	-0.871	0.605	2.071	0.150
Experience of GBV at childhood	0.194	0.325	0.359	0.549
Involvement to LAS interventions	-3.172	0.619	26.269	0.000
Awareness of women rights	-0.021	0.033	0.393	0.531
Religion	-0.264	0.329	0.643	0.423

Model fitting information final chi-square= (p < 0.000); Goodness of fit Deviance Chi-Square=328.967 (P=1.000) and Pearson’s Chi-square= 495.444 (0.167); Cox and Snell = 0.340, Nagelkerke = 0.409; Test of Parallel line = 0.363

3.6. Conclusions and Recommendations

This paper concludes that the greater proportion of women in Kongwa and Morogoro Rural districts had negative attitude towards GBV against women. With regard to involvement of women in LAS interventions, a number of non-beneficiaries had positive attitude towards GBV against women relative to beneficiaries. It is also concluded that ethnic background and women involvement in LAS interventions had strong effect and negative implication on women’s attitude towards GBV. It is further concluded that traditions dominated by a patriarchy system affirms men’s attitude towards GBV against women, while some women also uphold harmful traditional practice such as FGM.

In view of the conclusions based on findings from this paper, it is recommended that the government should mainstream LAS to all communities. LAS providers should put more efforts on sensitizing communities about existing wrong myths fuelling GBV against women; also sensitization on the availability and services offered by LAS providers is crucial for successful uptake of LAS. When planning for sensitization interventions, planners should consider ethnical backgrounds (specifically community with patrilineal background) by increasing education sessions on different forms of violence practiced against women and wrong myths which fuelling GBV against women. In addition, more studies to examine attitudes of men towards LAS interventions are recommended. This will generate more indicators to enable the planners and LAS providers to design and implement appropriate intervention programmes.

Funding: This study received no specific financial support.

Competing Interests: The authors declare that they have no competing interests.

Contributors/Acknowledgement: All authors contributed equally to the conception and design of the study.

REFERENCES

Abeid, M., P. Muganyizi, S. Masawe, R. Mpembeni, E. Darj and P. Axemo, 2015. Knowledge and Attitude towards rape and child sexual abuse – a community base cross section study in rural Tanzania. Biomed Central Public Health, 15(1):428.View at Google Scholar | View at Publisher

Agresti, A. and B. Finlay, 2009. Statistical methods for the social sciences. 4th Edn., New Jersey: Pearson Prentice Hall. pp: 609.

Ajzen, I., 1991. The theory of planned behaviour. Organizational Behaviour and Human Decision Processes, 50(2): 179-211.View at Google Scholar 

Anderson, C.L., 2005. Theorizing gender in intimate partner violence research. Sex Roles, 52(11-12): 853-865.View at Google Scholar | View at Publisher

Antai, D.E. and J.B. Antai, 2008. Attitude of women towards intimate partner violence: A study of rural women in Nigeria. Rural and Remote Health. Retrieved fromhttp://www.rrh.org.au/articles/subviewnew.asp?ArticleID=996 [Accessed 20/4/ 2015].

Arango, J.D., M. Morton, F. Gennari, S. Kiplesund and M. Ellsberg, 2014. Gender equality and development interventions to prevent or reduce violence against women and girls. A Systematic Review of Reviews. Retrieved from http://www.Worldbank.org/gender/agency [Accessed 20/4/2015].

Bailey, D.K., 1998. Methods of social science research. London: The Free Press Collier Macmillan Publisher. pp: 345.

Danish Institute for Human Rights (DHIR), 2011. Access to justice and legal aid in East Africa. A comparison of the legal aid schemes used in the region and the level of cooperation and coordination between the various actors. A Report by the Danish Institute for Rights, Based on Cooperation with East Africa Law Society. DHIR pp: 176.

De Vaus, D., 2002. Analyzing social science data. London: SAGE. pp: 401.

Ellsberge, M. and L. Heise, 2005. Researching violence against women. A practical guide for researchers and activist. Washington DC. United State: World Health Organization and Programme for Appropriate Technology in Health (PATH). pp: 259.

FEMNET, 2001. Training manual on gender based violence. Retrieved from .http://www.Mediterraneas.org [Accessed 20/7/2013].

Field, A., 2009. Discovering statistic using SPSS. 3rd Edn., London: SAGE Publications Limited. pp: 821.

Flood, M. and B. Pease, 2009. Factors influencing attitude to violence against women. Trauma, Violence, & Abuse, 10(2): 125-142.View at Google Scholar | View at Publisher 

Legal and Human Right Centre (LHRC), 2004. Legal aid provision in Tanzania improving access to justice. Legal and Human Right Centre. Dar es Salaam, Tanzania. pp: 37.

Legal Services Facility (LSF), 2012. Baseline survey on Tanzania mainland and Zanzibar for legal services facility. Retrieved from http://lsftz.org/publications/LSFBaselineSurveyReportFinal.pdf [Accessed 20/1/2013].

LHRC, 2012. Tanzania Human Rights Report. Legal and Human Right Centre. Dar es Salaam, Tanzania. pp: 436.

LHRC, 2013. Tanzania Human Rights Report. Legal and Human Right Centre. Dar es Salaam, Tanzania. pp: 436.

Likert, R., 1932. A technique for the measurement of attitudes. Archives of Psychology, 22(140): 1-55.View at Google Scholar 

Merry, S.E., 2006. Human rights and gender violence, translating international law into local justice. Chikago and London: The University of Chikago Press. pp: 269.

Nachar, N., 2008. The Mann- Whitney U:A Test for assessing whether two independent samples come from the same distribution. Tutorials in Quantitative Methods for Psychology, 4(1): 13 - 20.View at Google Scholar | View at Publisher

Nair, R., 2014. Violence against women, violation of human rights: An appraisal on Indian context. International Journal of Development Research, 4(11): 2357-2363.

Nayaki, M.B., C.A. Byrne, M.K. Martin and A.G. Abraham, 2003. Attitudes towards violence against women a cross- nation study. Sex Roles, 49(7/8): 333-342. View at Google Scholar 

NBS and ICF Macro, 2011. Tanzania demographic and health survey 2010. Dar es Salaam, Tanzania: NBS and ICF Macro. pp: 451.

Nyange, T.M., A.N. Sikira and J.G. Lyimo-Macha, 2016. Gender based violence and legal aid services interventions among rural women in Morogoro rural and Kongwa Districts, Tanzania. International Journal of Asian Social Science, 6(8): 446-461.View at Google Scholar | View at Publisher

Osborn, D., A. Cutter and F. Ullah, 2015. Universal sustainable development goals: Understanding the transformational challenge for developed countries. Stakeholders Forum. Retrieved from http://www.stakeholderforum.org [Accessed 19/10/2015].

Pallant, J., 2007. Statistical package for social sciences (SPSS) survival manual: A step by step guide to data analysis using SPSS for windows. 3rd Edin., Berkshire: Open University Press. pp: 133.

Sikira, A.N., 2010. Women empowerment and gender based violence in Serengeti District, Mara Region, Tanzania. Thesis for Award of PhD Degree at Sokoine University of Agriculture, Morogoro, Tanzania. pp: 207.

Sikira, A.N., E.A. Mwageni and R.J. Salanga, 2010. Attitude of women towards gender based violence in Serengeti District, Tanzania. Journal of Continuing Education and Extension, 3(2): 13-23.View at Google Scholar 

United Nation (UN), 1993. Declaration on the elimination of violence against women the general assembly resolution 48/104. Retrieved fromhttp://www.un.org/documents/ga/res/48/a48r104.htm [Accessed 1/1/2017].

United Republic of Tanzania (URT), 2013. 2012 population and housing census population distribution by administrative areas. Dar es Salaam: NBS. pp: 244.

World Health Organization (WHO) and Pan Africa Health Organization (PAHO), 2012. Understanding and addressing violence against women. Retrieved from http://apps.who.int/iris/bitstream/10665/77433/1/WHO_RHR_12.35_eng.pdf [Accessed 16/ 05/2016].

---