In thisstudyIRH isusing a social networkanalysis perspective. Network analysisacknowledgesthatindividualsdon’texist in a vacuum. People interactwith, learnfrom, and get information fromotherindividuals. And this type of analysisis more interested in theserelationships and the characteristics of thoserelationshipsratherthanindividual’sdemographic and socioeconomiccharacteristics. So in yes, an individual has demographic and socioeconomiccharacteristics, but network analysissaysthatrelationshipsbetween people and characteristics of thoserelationships are just as, or even more, important thanthoseindividualcharacteristics.
Demographers posit that fertility declines are the result, in whole or in part, of the diffusion of new knowledge and ideas from one locale, social group, or individual to another.7“Who delivers the message, and in what interpersonal context, may be just as, if not more important, than the message itself, and may result in better, more relevant, and perhaps more effective programs.”8Valent et al, 2006Women and men may receive accurate RH information from established health institutions; but they also make their decisions based on stories that circulate in their informal social networks, and they supplement provider instructions with informal conversations. Observing, discussing, criticizing, and evaluating, people pass information from one to another and from public sources to groups. Communication along interpersonal channels and through multiple media channels provide information about the existence of new behaviors, narrows uncertainties regarding the consequences of new choices and reduces the costs of innovation by modifying social norms. Individuals who are highly interconnected and centrally located within social networks are likely to hear about innovations earlier and have more opportunity to evaluate their benefits.9 Studies of the diffusion of FP information have found that having a direct or indirect link to the source of information was associated with either increased knowledge or use of contraceptives. In addition, the composition of an individual’s personal network and their position within the network has been associated with FP knowledge, attitudes, and use.
Social network analysis is best done using the ethnographic sandwich.The first ethnographic data collection, which we report on in another presentation, provides the context and language to use when collecting the census data. It consists of participant observation and in-depth interviews.The complete social network map of the community is the focus of this presentationThe second ethnographic study, which is ongoing, consists of in-depth interviews. It ensures that results collected in the earlier phases are correctly interpreted.
In this phase of the study enumerators nocked on all doors in the village, to create a complete list of all adult living the village. Individuals did not need to be at home to be recorded – any adult who was home at the time provided this based information about all adults living in the house/compound.
Ask: “What is your relationship with (name of person)? You can mention more than one relationship. For example, a person can be your aunt, and can also be your health provider.” Write the relationships. Relationships can be family relationship (such as mother, husband, sister, co-wife, etc.) or friend, or it can be a different type of relationship, such as shop-keeper, co-worker, health worker, teacher of the respondents children, member of her grin or tontine, etc. It can also be the maribout or religious leader.(c) Ask:”Does (name of person) live in this village, or elsewhere?” If ‘elsewhere’, ask “In what town does (name of person) live?(d) Ask:” How close would you say you are to (name of person)? Very close, close, or not close?” Write the response.(f) ASK: “Does this person influence your decisions about how many children to have and when to have them”?Ask: Would you say that (name of person) approves or disapproves of women using family planning to have fewer children? Mark Yes, No, or Don’t know.PROBES ARE FROM ETHNOGRAPHIC RESULTS
Younger women seem less connected, however no differences seen in education and parity
Koutiala, networks are large and dense. Everyone knows each other. Entering information circulates throughout entire network and thus reinforced (can be positive or negative information)Bandiagara, networks are smaller and community is less dense (fewer people know each other). Many people are named who live outside of the community. Entire network (including those not interviewed) Pink: womenBlue: menGreen: People who were named by respondents, but who were not interviewed (they are not married, too old, or live outside of the village)
Without the people circled in yellow, the three subgroups wouldn’t be connectedEntire network: influencers – BandiagaraHow does this network differ? Why do you think it is different?Are there any individuals more important than others? Who? What makes these people bridges?What do the yellow circles represent?What kind of information do they pass to each group? Is it different?
This map now is looking only at the network of men and women interviewed.Looking at Connectivity now- remember that means shortest path…to others – power brokers, gatekeepers.Orange is approvalGreen is disapproval.What do you see here? Who is more influential? How might this influence our program?
Look at the lightest color on the top of each bar- this represents the men and women who don’t know what their partners think.Especially look at women in the less favorable setting – majority don’t know – because don’t discuss.In Koutiala – only about 50% think partners are positiveIn Bandiagara- 40% of men and less than 20% of women
This network of women shows unmet need. The size of the circles = the amount of influence. The green circles are the users, while the read are women with unmet need. What do you see of interest here? How many users are there? Are there enough to bring out about a more favorable environment for FP?\\ How would you use snowball intervention to increase FP use. (User to user)
The power of relationships: Using social network theory and methodologies to improve family planning attitudes and use in Mali (ENG)
The power of relationships Using social network theory and methodologies to improve family planning attitudes and use in MaliNahyun Kate ChoIrit SinaiMariam DiakiteMamadou CamaraSusan IgrasRebecka Lundgren
2 Background 5 year project: Terikunda Jékulu • Phase 1 (one year): Formative research • Phase 2 (four years): Intervention and scaling up Project team: IRH (prime) CARE, CEDPA, ASDAP Based in Bamako • Formative research conducted in Bandiagara and Koutiala districts • Initial interventions to be carried out in Dire and Selingue districts
3 Social network analysis: What is it? A theoretical perspective applied to research and programs • Recognizes that individuals interact with, learn from, and get information from other people • Focuses on relationships, not individuals “Who delivers the message, and in what interpersonal context, may be just as, if not more important, than the message itself, and may result in better, more relevant, and perhaps more effective programs.” - Valente & Fosados, 2006
4 Objectives 1. Apply social network analysis approach to assess influence of social groups on: • Fertility attitudes, beliefs, desires, intentions and behaviors • Couple communication • Decision-making regarding timing & number of children • FP adoption & continuation • Method choice and switching
5 Objectives 2. Design and test interventions to activate key actors within a social network: • Reduce negative determinants • Strengthen positive influences on fertility attitudes and FP use 3. Expand interventions to additional communities (if proven successful)
6 How do we define unmet need? Number or percent of women currently married or in union who are fecund and who desire to either terminate or postpone childbearing, but who are not currently using a contraceptive method Do men have unmet need? Does it count as “met need” if the method used is unsafe or ineffective? When asked directly, do women think they have unmet need?
Formative research:The ethnographic sandwich Participant observation and in-depth interviews • Meaning & value of fertility-related communication • Distribution of social influence within community groups Ethnographic study Complete social network Community census map of the community Ethnographic study In-depth interviews with men and women identified from the census representing different unmet need and social network statuses
8 First step in mapping: Obtain a list of all adult men and women in the village, and assign each a code
9 2nd step in mapping: Identify from the list all women of reproductive age, and men married to women of reproductive age, and interview them Response rate: o 92 % in Koutiala, o 84% in Bandiagara Interview obtains background information, unmet need status, attitudes toward child spacing, and a social network module
Eliciting social networks, cont’dSocial network module For women, probe on: husband, mother, mother-in-law, co-wives For men, probe on: co-wives, father, male relatives
11 Analysis • Use SPSS to obtain descriptive statistics and cross tabulations, and identify respondent’s unmet need status • Use UCINET, a special software designed to analyze network data, to create village maps, and organize them by selected variables (e.g., sex, age, FP attitude and use)
12 Profile of Respondents in Each Community Koutiala Community Bandiagara Community More Favorable Less Favorable Environment Environment Women Men Women Men (n=217) (n=149) (n=208) (n=152)Age (mean) 27.3 39.5 Age (mean) 30.9 40.0No. children 3.3 5.0 No. children 3.8 5.7(mean) (mean)Formal education 95 97 Formal 94 87(% no education) educationReligion 84 90 (% no education)(% Moslem) Religion 100 99Marital status (% Moslem)(% polygamous) 54 37 Marital status (% polygamous) 48 32 12
Unmet Need Status of Respondents Koutiala Community – More Bandiagara Community – Less Favorable Environment Favorable Environment Women Men Women Men (n=217) (n=149) (n=208) (n=152)Pregnant 21% 27% Pregnant 15% 1%Desires child 17% 30% Desires child 16% 22%this year this yearUses modern 12% 9% Uses modern 9% 7%method methodUses Traditional 2% 7% Uses Traditional .5% 2%method methodBF or PP 17% 8% BF or PP 19% 15%amenorrhea amenorrheaInfertile/no sex/ 2% 11% Infertile/no sex/ 9% 3%menopause menopauseUnmet need 30% 35% Unmet need 32% 34%
14 Connectors and Influencers Influencers are the most nominated individuals in a network. They have direct influence on more people. A high score indicates the opportunity to influence and be influenced directly. Connectors are the individuals with the shortest path to all other nodes They are gatekeepers, brokers, control the flow, liaise between cliques. A high score Indicates power and access, but also could be a bottleneck. 14
How are the social network mapsorganized? Social Network View Outcomesnetwork status• Connector • Entire • Community• Influencer network structure • Men & • FP attitudes women • Unmet need • Men only status • Women only 15
16 Does womens network status vary by age? Koutiala Bandiagara More favorable Less favorable Influencer 30* 32.7 (mean age) Network Connector 30.3* 31.3 Status (mean age) Neither 26.1* 29.9 (mean age) * P < .01
17 Social network maps of villages (entire network) Koutiala Bandiagara Higher unmet need Lower unmet need More favorable Less favorable Women Men Not interviewed Size = Influence
18 Entire social network in one village in Bandiagara: Influence Womens Mens NominatedSize = Influence
Positive FP attitudes among men and women in Koutiala : Connectivity Approval Disapproval Women MenSize = Connectivity
Do your network partners approve of family planning? Koutiala/More favorable Bandiagara/Less favorable120 100100 8080 6060 404020 20 0 0 Men Women Men Women Yes No Dont know Yes No Dont know
Unmet need among women in one village of Bandiagara: Influence Unmet need FP use Met needSize = Influence
Key results: Social barriers• Perceived spousal and community disapproval of FP• Stigma related to talking publicly about FP, little discussion• Misinformation about FP• Women unable to communicate or make decisions about child spacing with partner• Disapproval of FP: religious reasons and because it is a negative, western concept
Key results: Social networks• Women receive material support from husbands and brothers-in-law, practical support from sisters-in-law and co-wives and cognitive support from mothers-in-law• Men not well-informed, nor linked into discussions on FP• Men and women unaware of what friends and sexual partners feel about FP
24 Social network findings (cont.) • Co-wives, mothers and mothers-in-law are motivated to support FP use because they would be responsible for looking after the children. • Rumors and misinformation are spread through networks ad represent a key barrier to FP use. • Men and women are not aware of their spouse’s view of FP. • Providers not linked into social networks
25 Programmatic implications • Working with and through women’s groups • Working with and through men’s groups to provide information and dialogue • Engaging community- based religions leaders • Bringing home new ideas through migrant workers