Statistical Inference & Network Models
ECI 589 SNA and Education: Unit 4 Readings
Overview
In Unit 4 we shift our focus from mathematical approaches for describing networks to statistical approaches for explaining and ultimately predicting network properties and outcomes for those embedded within these networks. A secondary goal of readings and discussion is to help you generate ideas for independent application of network analysis. As part of our readings, for example, you’ll be introduced to a range of questions that statistical models for network data can help answer.
Readings
For Unit 4, you will read chapters 8 and 9 from Social Network Analysis and Education (Carolan 2014) and locate one additional research article, presentation, or informal study to help address our discussion questions for the week.
SNA and Education Part III: Applications and Examples
The following readings move us beyond techniques introduced in Parts I & II for describing networks and focuses on recent advances in inferential statistics that can be used to make predictions from social network data and test hypotheses we have about a network of interest. We’ll learn about different techniques that make use of simulations to model network data and how these statistical models are used to address questions that more completely reflect the complexity of educational settings.
Self-Selected Study
Use the NCSU Library, Google Scholar or search engine to locate a research or practitioner-focused article, presentation, or resource that applies social network analysis to an educational context or topic of interest. More specifically, locate an network study that makes use of statistical models to explain or predict an education related process or outcome. You are also welcome to select one of the research papers listed in the syllabus or in our course text that may have peaked your interest. Your selection should address one or more of the discussion topics/questions (see below) and you’re welcome to identify SNA resources outside of the field of education.
Discussion
In lieu of the peer interaction and discussion of course materials that normally take place “in-class”, you’ll be asked to log in this week and engage with other members of our learning community through this discussion forum. With the exception of the Self-Selected Study questions, you are not required to address every guiding question, particularly if you feel others in the class have thoroughly addressed the topic or questions. Our aim for these discussions is to collectively build our understanding of these readings through back-and-forth dialogue and avoid a “collective monologue” in which we see 20 variations of effectively the same post.
To create a new post in response to one or more of the guiding questions listed below, click on the forum post associated with the chapter questions and then click “Reply” to add your response. Remember, you are not required to address every guiding question with the exception of the Self-Selected Study questions for which you must create one new discussion topic and respond to the discussion prompt and required questions.
Guiding Questions
To help guide our discussions, we will collectively address a set of guiding questions provided below. You are also welcome to add your own topics or questions for the class to discuss.
Self-Selected Study
For your self-selected study, provide an APA citation and abstract, then briefly answer the following questions as appropriate:
What is being modeled: ties in complete networks, individual actor attributes, or groups of actors?
What procedure was used and what, if any, covariates were included in the models? At what level were these covariates measured: continuous (ratio or interval) or categorical?
Did the modeling strategy measure a network’s change over time?
How, if at all, were simulations used to create a probability distribution against which observed network parameters were compared?
What does it mean when the result from a statistical model using network data is reported as being “statistically significant”? How does this relate to the implied null hypothesis?
Chapter 8: Statistical Inference With Network Data
Answer one or more of the following questions:
- Why are simulations necessary in order to make probabilistic inferences with network data?
- Explain in plain language how simulations are used to create a probability distribution that enables you to make a statistical inference with network data.
- Contrast the aims of the mathematical and statistical approaches to social network analysis. For what reasons would educational researchers prefer one approach versus the other?
Chapter 9: Network Data and Statistical Models
Using one of the studies mentioned throughout this chapter, identify the statistical model that was employed and answer one or more the following questions:
Do you think the choice of statistical model(s) used was appropriate?
If this same study were to use standard statistical models that assume independence among observations, how would this influence the study’s results?
Assume you had network data from an entire high school student body (N = 250) and were interested in predicting a student’s number of friends from covariates such as sex, grade level, and academic performance. What model would be most appropriate to test these relationships?
Assessment
Grading for this assignment is fairly lenient, provided that it’s clear from your posts that you’ve done the required reading. Readings and discussion for each unit are worth 6 points and judged based on quantity and quality of your posts.
In term of quantity (3 points), you’ll be expected to create a new discussion topic your self-selected study and add at least 3 new posts and/or replies for a total of 4 posts. So others will have an opportunity to read and respond to your posts, your posts should be spread out over the course of the week and across at least two different days, preferably not the last two days.
In terms of quality (3 points), your posts over the next week should provide new or insightful contributions to that question or topic. There is no requisite for the length of each posting, in fact short conversational exchanges (1-3 paragraphs) are highly encouraged. I strongly recommend looking at the Productive Online Discussion Model copied below from Gao, Wang, and Sun (2009) for ways to contribute to the conversation.
At minimum, your collective posts should also help us interpret or elaborate on discussion topics, questions, or ideas others have shared by “making connection to the learning materials” and should reference at some point each of the required chapters and your self-selected reading or resource.
Productive Online Discussion Model
Disposition 1: Discuss to Comprehend
Actively engage in such cognitive processes as interpretation, elaboration, making connections to prior knowledge.
- Interpreting or elaborating the ideas by making connection to the learning materials
- Interpreting or elaborating the ideas by making connection to personal experience
- Interpreting or elaborating the ideas by making connection to other ideas, sources, or references
Disposition 2: Discuss to Critique
Carefully examine other people’s views, and be sensitive and analytical to conflicting views.
- Building or adding new insights or ideas to others’ posts
- Challenging ideas in the texts
- Challenging ideas in others’ posts
Disposition 3: Discuss to Construct Knowledge
Actively negotiate meanings, and be ready to reconsider, refine and sometimes revise their thinking.
- Comparing views from the texts or others’ posts
- Facilitating thinking and discussions by raising questions
- Refining and revising one’s own view based on the texts or others’ posts