Why would you use NETWORKS 3 or 6 or NETJAWS (and not UCINET)?
- You have observation data (SCM: people report about peer groups, they don’t just provide reports about themselves, but also reports about everybody else)
- You don’t trust everybody’s report the same way (e.g., your data are on children, who may show self-enhancement biases) and only want to accept ties you can trust above chance
- You don’t mind highly interconnected, fuzzy, and overlapping networks, but you do want everybody’s connections to everybody else
- You want to determine reliability of your composite map (and of individual reporters)
- You want reliability indices for people’s ties
- You just want a co-occurrence matrix for use with other strategies (e.g., to import into SPSS for factor analyses; to import into UCINET)