X hits on this document

PDF document

The author(s) shown below used Federal funds provided by the U.S. Department of Justice and prepared ... - page 28 / 48





28 / 48

Columbia’s CCP: A Case Study

work.3 The CCP proposals (prepared by the sites) were used by the research team to identify a preliminary list of potential actors and organizations within the CCP network. These lists were mailed to the CCP project director for review, who then recommended deletions and additions. The realist ap- proach uses the criterion of “mutual relevance” to decide who belongs in a network. Here, the assumption is that individuals and groups are included in the network if they have a mutual interest in the CCP project and some ca- pacity to influence the outcome. Indeed, there is reason to believe that indi- viduals were included in the proposal (or later included in the network) be- cause of their position in particular organizations or projects associated with CCP.

Sampling was not necessary in this study because the network populations were relatively small. Hence, all identified members of each network were included in the data collection effort.

Data Collection Methods and Procedures

The network data in this case study were collected as part of our Coalition Survey. The Coalition Survey was sent to sites from September, 1995 to June, 1996, depending on the site. This network analysis then is a snapshot of the relationships and social networks during the first half of the CCP im- plementation phase.

To measure CCP-related networks, respondents were given a list of individu- als who were believed to be affiliated with the CCP coalition in their respec- tive cities, and then asked how often they have contact with each individual on the list. Possible response options were “daily, weekly, monthly, every few months, never.”

To enhance the network analysis, individual cases were dropped when they did not have sufficient contact with other members of the network. Including persons with rare or occasional contacts in the network would have distorted the results by causing more dense (and therefore less interpretable) cluster- ing of the remaining actors. Hence, after examining the frequency distribu- tions, a decision was made to include only respondents who reported having contact with at least 10% of the total network “at least every few months.”

The effects each site.


applying this inclusion criterion are described The analysis strategy can be found in





3The realist approach can be contrasted with the nominalist view. With the latter, network boundaries are determined by the researcher’s theoretical framework.



Document info
Document views63
Page views63
Page last viewedSun Oct 23 14:46:49 UTC 2016