This course emphasizes a current focus in learning analytics on the importance of social media and social networking. One hypothesis is that students commit to learning and learn in part through their interactions with other students and with faculty members.

It is clear to me that there should be a correlation between being involved in a class and learning. If I spend more time with the class I’ll probably develop more connections with other students, post more on message boards, and also get better grades.

But is this relationship causal?  That is, will students learn more if they interact with other students than they would if they did not?  I’m more sceptical about this possibility and I’d like to know if it is true.

A paper published by one of the instructors in this course , Dr. Dragan Gasevic, presents some examples of this relationship:

  • Johnson and Johnson (2009) talk about the importance of cooperative learning, arguing that this approach is has “progressed to being one of the dominant instructional practices throughout the world.”  The idea of “social interdependence” which is the foundation of this approach owes debts to Kurt Lewin and Morton Deutsch.
  • Lev Vygotsky, the Russian psychologist who argued that children’s involvement with knowledgeable community or family members helps them learn to think and behave in ways appropriate for their society.
  • Vincent Tinto argues that the importance of student integration into the learning community is a critical determinant of whether a student is going to complete college or not.

In his paper the researchers measure whether students have taken classes together over time in an online learning program and use these measures of networking to predict grades. The networking variables predict learning. This seems consistent with the correlational explanation.

But it’s not so clear to me why this happens. Is spending time with the same students a causal factor?

Dr. Gasveic, also recommended these readings:

I also found this article, which I found to be a gem because it attempts to assess the causal relationships among people and grades .

http://theunbrokenwindow.com/Higher%20Ed/Readings/zimmerman%20peer%20effects.pdf

I particularly love this:

“The production of higher education is characterized by an unusual customer input technology whereby student quality is arguably a key input into the production of educational services—students may learn better when in the company of other strong students. The fact that students are themselves the only providers of this potentially key input in the production of education could explain why schools care so much about the characteristics of their customers and why elite schools create a queue of applicants from which to select by regularly setting their tuition well below the full cost of the education they provide (cf. Winston, 1996).

… and also that he dares cite this wonderful book, which is hated by many developmental psychologists:

Harris, Judith, The Nurture Assumption, New York: The Free Press (1998)

The effects are small, as we might expect. But this is a great approach to assessing them.

One of our classmates, Cecilia, posted this regarding the Zimmerman paper:

I loved this part: “”….mixing rather than segregating students of different abilities may generate higher aggregate learning. Intuitively, the increase in learning from moving a weak student to a peer-rich environment exceeds the loss in learning from moving a strong student to a peer-poor environment. This logic parallels the justification for income equalization in a world with diminishing marginal utility of income.” (pg.10)

Social Network Analysis