The “Small World Phenomenon” : How “Small” is a University?
The Small World Experiment
We learned in class about Stanley Milgram, who in the 1960’s, conducted an experiment with other researchers to find out the average path length for social networks in the United States. In his experiment, he had a group of people in Nebraska send out letters that, hopefully, would eventually reach his target, a stockbroker in Boston, Massachusetts. In his experiment, only 64 of the 300 letters were received by the target and from the results, Milgram found out that the median number of steps is 6, which later on, will be correlated to “six-degrees of separation” and the “small world phenomenon.”
Milgram’s experiment and results were groundbreaking and it eventually lead to further research into just how “small” the world is.
Other “Small World” Experiments
http://www.analytictech.com/connections/v20(2)/smallworld.htm
In the above article titled “Small World in a University,” an experiment was conducted into seeing just how “small” a university is. In this experiment, 60 students were randomly selected from the university where 15 were freshmen, 15 were sophomores, 15 were juniors, and 15 were seniors. The target was the undergraduate dean of the management school. Each student was given a folder with forms inside where the students and eventually, the people they past the folder onto, will sign their name and personal information (so that there will be no backward passing). Students were only allowed to pass the folder to someone whom they have spoken to multiple times outside of a classroom environment, so that although the tie may not be necessarily strong but it is not weak enough that it is just “passing onto the student next to you.” Thus, all folders were passed from person to person.
An interesting point in this article was that there were several hypotheses that this experiment also tested. The four hypotheses were:
(1) The longer the time at the university, the more likely a student is to initiate a successful chain of communication to a target.
(2) Small world folders are more likely to be passed within a class than between classes and occupational groups in a university.
(3) Small world folders will converge on faculty and staff before reaching the target.
(4) Small world folders are more likely to be passed to members of the same sex.
(The reasoning behind these hypotheses is further explained in the article.)
Results
By the end of the experiment, 16 of the 60 folders eventually reached the target, which is similar in percentage to other past Small World Experiments. There was a mean of 1.25 links between the students and the target. According to the article, this result is quite plausible as the university is a much smaller network compared to the network in Milgram’s experiment. It is very likely that within an organization like an university, the links between starter and target is much smaller.
The results for the hypotheses are as follows:
(1) For hypothesis 1, 81 % of the 16 (successful) folders started from a sophomore or junior but only 2% were from a senior. The expectations that seniors will be more successful does not necessarily seem true. However, most of the chains consisted of passing to upper-class students rather than lower class students especially if it started from an upper-class student.
(2) Hypothesis 2 seems to be true – students are more likely to pass folders within classes and not outside even though passing to someone outside will have a greater chance of success.
(3) There were no indications as to whether hypothesis 3 is true – the links were not long and it does not appear that there was a convergence. However, graduate students, staff and faculty does seem to be more closely tied to the administration where the target was in.
(4) From the results, it seems that females are more likely to pass the folder to other females than males were to pass to other males.
(The author does note that because the experiment’s size was so small, the conclusions and results may not accurately reflect the true network in an university.)
Conclusion
Milgram’s experiment has lead to many other “small world” experiments that try to duplicate Milgram’s results. Of course, there are other “small world” experiments such as the one in the article where the experiment was conducted in a smaller network. In the article that I read the mean was 1.25 links while in Milgram’s experiment, the number of steps was approximately 6. (Of course, a university is a much smaller network). It was interesting to see that other information can be taken from the results of these experiments that gives information regarding the network (e.g. the four hypotheses). It would be interesting in the future to see further “small world” experiments (perhaps at Cornell?) and their results.