Section 3: Social Cohesion: Communication for Social Change: An Integrated Model for Measuring the Process and Its Outcomes

Social Change Outcome Indicators

Social Cohesion

Social cohesion consists of the forces that act on members of a group or community to remain in, and actively contribute to, the group. In cohesive groups, members want to be part of the group, they generally like one another and get along well, and are loyal and united in the pursuit of group goals. Social cohesion is an important antecedent and consequence of successful collective action. Social cohesion mediates group formation, maintenance, and productivity.

Dimensions of Social Cohesiveness

For the purposes of this study, social cohesion can be divided into at least six related social and cognitive dimensions:

• Sense of belonging,
• Feelings of morale,
• Goal consensus,
• Trust,
• Reciprocity, and
• Network cohesion.

Sense of belonging:

Is the extent to which individual members feel as if they are an important part of the group or community. The group's level of belonging can be measured by means of focus group discussions, in-depth interviews with individual members and sample surveys of community members. The following items may be used for this purpose; responses should be coded as, strongly agree, agree, unsure, disagree, strongly disagree (standard Likert-type format):

1. I feel that I belong to this community.
2. I see myself as part of this community.
3. I feel that I am a member of this community.
4. I would rather live in a different community/village.
5. I would rather live in this community than any others I know of.
6. I would like to move out of this village as soon as possible.
7. People in this community are all striving for the same goals.
8. Everyone here wants to pursue their own goals rather than working for the good of the community.

Feelings of morale:

Refer to the extent to which members of a group or community are happy and proud of being a member. Level of belongingness can be measured by means of focus group discussions, in-depth interviews with individual members and sample surveys of community members. The following items can be used for this purpose; responses should be coded as, strongly agree, agree, unsure, disagree, strongly disagree (standard Likert-type format):

1. I am happy to be part of this community.
2. I am content to be part of this community.
3. This community is one of the best anywhere.
4. I want to work with the same people on our next community project.
5. I would rather work with different people on our next community project.
6. Most of the people in this community project genuinely like one another.
7. Most of the people here are willing to share responsibility for making our community a better place to live.
8. There are too many people in this community who think they should share in the benefits without contributing their share of the work.

Goal Consensus:

Goal consensus is the degree to which members of the community agree (1) on the importance of each problem or issue facing the community, and (2) on the objectives to be achieved by the group. Agreement assumes shared knowledge. Before members of a community can reach an agreement regarding priorities and goals, they must know what the issues are and what objectives have been set. (See, the convergence model of communication described in the first section of the document.) During the initial stage of investigation when the focus group discussion's are being held to define the community, its organisations and leaders, they should also be used to elicit the list of the 10 most important problems/issues that the community faces. This consensus list of problems can then be used in a survey of the remaining community members to measure goal consensus. Each respondent is presented with a list of these 10 problems/issues and then asked to indicate to the interviewer which problem is the most important to them at this time, which is the second most important issue, and so forth, for the entire list.

Indicator:

Each of the 10 problems will have a score (rank) from each respondent in the survey that ranges from 1 to 10. If a problem is judged to be the most important priority, then it would receive a score of 1; the problem with the lowest priority would receive a score of 10. For each problem, one can compute an average rank score, which is the sum of scores given by each individual divided by the number of individuals in the sample.[4] The problem with the lowest average rank is considered by the community to be the most important problem. One can also assess the agreement on this ranking by calculating the variance around this average rank.[5] For example, if every person in a survey of, say, 30 individuals judged water purity to be the No.1 priority, the average rank will be the sum of all the 1.0 (equal 30) divided by the number of individuals (30), which gives a mean score for water purity of 1.0. The difference between each person's own score and the average rank would be zero if everyone gave it a score of 1.0, as in the example. In this case, the average variance would be 0, indicating perfect consensus on that problem. If everyone agreed that malaria was the second most-important problem, then the average rank score for the community would be 2.0, and once again the variance would be 0, indicating a perfect consensus on its rank order. Perfect agreement on the rank of all 10 problems would yield variances of 0. Maximum disagreement would occur when half the community ranked a problem as first priority (1.0), and the other half ranked the same problem as last (10). Dialogue about the community's priorities is expected to reduce the variance or disagreement on these priorities, and produce a greater, but not necessarily perfect consensus. The calculation of the actual variance in priority ranking scores can be used to know exactly how much progress has been made in reaching a community consensus on each of the 10 problems, after a process of community dialogue.

Social trust:

Is the general confidence that one has in the integrity, ability and good character of other people? Trust is sometimes thought of as the glue that holds a group or community together and makes cooperative action possible. Since 1972, the General Social Survey of the National Opinion Research Center in the United States has included a general question about whether or not other people can be trusted. Key informants of a community and/or a random sample of community members may be asked:

General:

1. Can other people (in this group/community) be trusted [check one]?

Can be trusted.

Cannot be trusted.

Not applicable.

Don't know.

Project specific:

Responses should be coded as, strongly agree, agree, unsure, disagree, strongly disagree (standard Likert-type format).

2. I do not trust others to have any influence over issues that are important to our projects.

3. I am comfortable giving other people responsibility for project tasks even when I cannot monitor what they do.

4. I can rely on the people that I work with on this project.

5. People in this group/community have confidence in one another.

Social reciprocity:

Refers to mutual interchange of favors, privileges and benefits in a relationship. For example, if someone helps another build their well or bring in their crop, the person who receives the favor is expected and actually returns the favor at a later date. Level of perceived reciprocity can be measured by means of focus-group discussions, in-depth interviews with individual members and sample surveys of community members. The following items can be used for this purpose; responses should be coded as, strongly agree, agree, unsure, disagree, strongly disagree (standard Likert-type format):

1. People behave in an opportunistic way and disregard their obligations to others.

2. People fulfill their obligations when they can be punished for not doing so.

3. People fulfill their obligations with others because, if they find out, the people around them will know that they are not trustworthy.

4. People in general fulfill their obligations to be sure that others will do so for them.

5. People try to fulfill their obligations to others.

Network cohesion:

Exists in a "bounded network or clique with a high level of internal cohesion, usually defined in terms of the density or connectedness of the links of information exchange among its members" (Kincaid, 1993, p. 113). A social network consists of all of the dyads or pairs of individuals (or groups) within a community that are linked by some form of social relationship (kinship, friendship, economic tie, etc.), while a communication network consists of all of the dyads or pairs of individuals (or groups) within a community that are linked by information exchange. Communication network data is obtained by means of personal interviews with all members of an intact network, group, or community and by asking them to name all (or some limited number, five to 10) members of the community with whom:

1. They have talked to most often in the last ___ (months/days) [General Interpersonal Network].

2. They have discussed [health problem x, y, or z ] in the last ___ months/days [Content Specific Network].

3. They have sought (or given) advice to about [health problem x, y, or z] in the last ___ months/days [Opinion Leadership Network].

Measures of network cohesion may be estimated:

1. By computing the density of a network of all individuals in a group or community, calculated as the number of pairs of individuals in a network that are linked to one another, divided by the total number of possible links in the network (see, Figures 3 and 4).

2. By computing centrality of a network based on the distance of all individuals to one another (number of steps of separation), calculated as the average number of the shortest steps connecting each individual to all others in a network (see, Figures 3 and 4).

[5] Using the formula for variance: s2 = (xi à √)2/n. Each individual's rank score for a particular problem, is subtracted from the average, then all the individual difference scores are added up and divided by the number of individuals in the sample to produce an "average" difference from the average rank score.