Network Analysis Resources
Welcome to the Berea College Hutchins Library network analysis resource guide.
What is network analysis?
Network analysis surveys the relationship between different entities, such as collaboration between researchers, interactions between genes, or communications between a people in a company. It can be utilized for a variety of purposes, from simply studying the structure of a community to solving complex math and engineering problems through graph theory. Along with this, network analysis can also examine the relationships between publications based on authorship, citations, standard terms, the spread of information, and even memes!
A network is simply several points (or ‘nodes’) that are connected by links. Generally, in social network analysis, the nodes are people, and the links are any social connection between them – for example, friendship, marital/family ties, or financial ties.
Types of networks: (Halgin & Dejordy 2008)
- Socio-centric used when analyzing the different patterns of interaction within a defined group.
- Egocentric is used when a research question examines a phenomena affecting individual entities across different settings.
Network visualization is the visual component of network analysis. There is a wide range of network visualization to choose from depending on the kind of data you have available or what types of relationships you want to see and show.
The network analysis process requires:
- Defining your research question
- Construct an academically driven question that refers to a connection of entities in some fashion.
- Ex. 1. I want to know how social relationships contribute to the construction and maintenance of an individual's nutritional health.
- Ex. 2. I want to analyze how restrictive censorship on Twitter has contributed to the rising pattern of equitable language on the social network site.
- EX. 3. I want to explore the amount of flow of information that was available to the Hong Kong protesters on social platforms in 2019.
- Identifying the kind of data needed for your analysis
- Who, what, and how much data is available to the information that I want to survey?
- Relational data: revealing some kind of connection between individuals, institutions, or products.
- Co-occurrence (same organization, same school, etc.)
- Distance (number of miles between, etc.)
- Actions (talk with, meet with, collaborate with, eat with, etc.)
- Resource (knowledge, facility access, resource access, etc.)
- Affective (like, dislike, respect, etc.)
- Kinship (e.g., sister, brother, cousin, etc.)
- Social roles (supervisor, teacher, friend, acquaintance, etc.)
- Existing data: public datasets on organizational connections, data on social media connections, datasets from CRMs (like Salesforce products (commodities, stocks, currencies), and your own existing knowledge on the matter.
- Binary Data:
- Interval Data:
- Select data collecting tools
- A survey should include questions regarding the background of the respondent and a way for them to provide information on connections.
- Select data collecting methods
- Full Network Method: used when collecting data from every member of your network (or network subset that you are investigating). This method works with a bounded network. You may not be able to get everyone, but the more people you get, the more complete your understanding of the network will be.
- Snowball Method: used when starting with a core group of network members, you collect data on all of their connections. Then you reach out to the new connections and collect data on all of their connections. This continues until you cannot surface any more new members or until you run out of time. This method will miss members who are not connected to the people sampled and may bias your sample; on the other hand, it may also help you access a wider sample of network members than you could have identified on your own.
-
Analyze data