innovation, organizations, and society

I attended a really interesting conference over the weekend co-sponsored by Northwestern and the UofChicago Booth School about innovation, organizations, and society. The conference included a diverse set of papers, the topics of which ranged from culture and cognition to network contagion to integrative capabilities of firms. The conference, organized by Pablo Boczkowski and Steve Kahl, left the impression that research on innovation and technology in organizations is alive and well. The entire set of paper abstracts can be found here.

Some highlights that resonate with some issues we’ve discussed here before – Sinan Aral‘s paper distinguished between the effects of homophily (i.e., the tendency of similar people to know each other and, consequently, to do the same kinds of stuff) and direct network links on the adoption of a technology. The paper very effectively shows that when you control for selection effects due to homophily, direct network links do not account for as much technology adoption as you might think. In fact, according to his figures (which are not available anywhere online) the effect of network links may be cut by more than half once you account for selection effects. The study has important implications for those who argue that network studies may overstate the effects of network ties on the diffusion of everything from obesity to happiness (see posts here and here).

James Evans presented a paper looking at the effects of online access to citation papers in science. The general hypothesis often put forward by Web evangelists is that the Internet makes people more likely to search far and wide for knowledge and thus flattens the information hierarchy.  Interdisciplinary citations should thus become more common when journals become open access. Evans finds that, while it’s true that people are more likely to cite research outside their own discipline, the focus of citations tends to be on a few central hubs. The Matthew effect in citation patterns doesn’t disappear, it gets bigger! Thus, this is more evidence that the long tail hypothesis is essentially wrong. The internet, despite its potential for linking people to the far ends of the world, tends to focus people on the same kinds of “popular” knowledge that everyone else is citing.


Written by brayden king

October 5, 2009 at 3:18 pm

Posted in brayden, research

6 Responses

Subscribe to comments with RSS.

  1. Is a draft of Aral’s paper posted anywhere?

    Also, I generally dislike the hype over the long tail, and find the concept misapplied everywhere, but preferential attachment like Evans describes is far from inconsistent with a long tail hypothesis. In a debate in the ’60s, Simon and Mandelbrot basically hashed this out, and the answer is that preferential attachment is one (but not the only) mechanism whereby long-tail networks can form.


    Michael F. Martin

    October 5, 2009 at 3:53 pm

  2. Dear All,

    This paper is currently under review at a journal that has strict press and website embargoes against early distribution. If you email me I am happy to send it to you… We are hope to have it available very soon.




    Sinan Aral

    October 5, 2009 at 5:48 pm

  3. Re the long tail: folks may find Tucker and Zhang’s paper ( to be interesting.



    October 6, 2009 at 2:20 pm

  4. Interesting! (Brayden sent me a link to this like a week ago, but I finally finished my last overdue thing so I feel like I can indulge). One of my favorite (old) papers showing the deflation of so-called “network-effects” once you control for homophily is this one.



    October 20, 2009 at 4:59 pm

  5. Thanks for the link Omar. The Cohen piece was published in 1983, which shows that network scholars have been aware of this problem for a while now.



    October 20, 2009 at 5:12 pm

  6. Grrrr…. tried to find the abstracts form conference, but the site has been taken down, I guess. Innovation without digital memory?



    March 28, 2011 at 4:09 pm

Comments are closed.

%d bloggers like this: