Скачать презентацию Network Analysis Applications for Qualitative Content Analysis Data Скачать презентацию Network Analysis Applications for Qualitative Content Analysis Data

6ad5db02c42466d2080bd383f0f9898d.ppt

  • Количество слайдов: 13

Network Analysis Applications for Qualitative Content Analysis Data Network Analysis Applications for Qualitative Content Analysis Data

What is Network Analysis? Ø A way of thinking sociologically Ø A body of What is Network Analysis? Ø A way of thinking sociologically Ø A body of sociological theory Ø A body of research findings Ø A form of quantitative-qualitative analysis Ø A set of analytical tools for analysis

A Different Way of Thinking Ø Approach is holistic—whole networks Ø Focus is on A Different Way of Thinking Ø Approach is holistic—whole networks Ø Focus is on relations between units Ø Organize data differently Ø Ask different questions l about the overall state of the network l about the nature of particular relations Ø Use different quantitative measures

Two Types of Network Analysis Ø One kind theorizes about social structure l l Two Types of Network Analysis Ø One kind theorizes about social structure l l uses network data to test hypotheses also uses simulated data to test hypotheses Ø The other kind is exploratory l l uses real data to figure out what is going on often uses visual network diagrams Ø Exploratory approach flexible, accessible

Two Kinds of Data Arrays Bob Carol Ted Alice Age 25 21 30 24 Two Kinds of Data Arrays Bob Carol Ted Alice Age 25 21 30 24 Gender Etc. M F Question: pattern of relations within the network Bob Question: pattern of characteristics in the sample Carol Ted Alice Bob --Carol 1 1 -- 1 0 0 1 Ted Alice 1 0 -0 1 -- 0 1

Some Possible Patterns Some Possible Patterns

Network Language Ø NODES---the individual units l l people groups nations companies Ø TIES—the Network Language Ø NODES---the individual units l l people groups nations companies Ø TIES—the relations between the units l l l Can be present/absent (Are X and Y linked? ) Can be directional (Does X like Y? ) Can be quantified (# of links, amount of trade)

Nature of Network Data Ø Populations, not samples preferred l l Do not use Nature of Network Data Ø Populations, not samples preferred l l Do not use probabilistic statistics Use matrix algebra as mathematical base Ø Ego-centric Networks—all ties from one node l l Only direct ties to ego Trace ties of those linked to ego Ø Full networks—all ties between all nodes l l can limit network to make this feasible can some times use snowball to trace network

Network Diagrams Network Diagrams

Some Network Measures Ø Density Ø Connectivity Ø Reachability Ø Distance Ø Reciprocity Ø Some Network Measures Ø Density Ø Connectivity Ø Reachability Ø Distance Ø Reciprocity Ø Clustering Ø Hierarchy Ø Cliques

Using with Content Analysis Ø Your data may include “relations” already l l l Using with Content Analysis Ø Your data may include “relations” already l l l website links or reports in the data ties between people, groups, or other units “grammatical” relations Ø You may be able to create relational data l l construct it from your Access database add external data or more codes Ø You might just want to diagram relations

Form of Network Data Enter directly in the network programs, UCINET, PAJEK, or NETDRAW Form of Network Data Enter directly in the network programs, UCINET, PAJEK, or NETDRAW (included in UCINET) Ø Import from an Excel spreadsheet (flat file matrix) Ø l l Ø Move data from Access to Excel Then import into UCINET Enter usual data array with other variables l l UCINET can turn it into a matrix for you You then can add the relations to the matrix Output a text file and format it for UCINET Ø UCINET can export into PAJEK format or Netdraw Ø

Network Resources Ø UCINET software: (free trial, $25 for students) <http: //www. analytictech. com> Network Resources Ø UCINET software: (free trial, $25 for students) Hanneman, Robert A. and Mark Riddle. 2005. Introduction to social network methods. Riverside, CA: University of California, Riverside ( Free tutorial, published in digital form at http: //faculty. ucr. edu/~hanneman/ ) Ø Pajek: network analysis and visualization: http: //vlado. fmf. uni-lj. si/pub/networks/pajek/ (free Ø software, buy better manual)