I Have no Dog in this Fight:

Reply to Blass' Comment on "Could it be a big world after all?":


Judith S. Kleinfeld

University of Alaska Fairbanks


I am very grateful to Professor Blass for his help on earlier versions of the "Small World Problem."  Stanley Milgram was a hero of mine when I was in graduate school at Harvard.  I tried to replicate his study (by postal and by e-mail) with high hopes. When I found Milgram's first unpublished study in the Yale archives showing 95 percent of the letters did not go through and yet realized he had quoted at length in his famous "Psychology Today" piece the anecdote about the one letter that did get through from Kansas to Cambridge in four days with only two intermediaries without reporting the dismal statistical results, I was very disappointed with my graduate school hero.


Well, we all grow up and find our heroes have their faults, and I still adore Stanley Milgram's verve and creativity and drama.  I tried to find the original subjects of his study, in fact, in order to replicate my own study with them.  I did find their names (you can find their names too in the archives) but I couldn't find them or their children.  I tried everything short of a paid detective service.  It was not a small, small world, I sadly concluded.


My paper is hardly intended to be the "last word" on the subject.  My paper is merely intended to be a cautionary tale.  I also hoped to recover the "small world" problem for psychologists, who might be curious as to why people are so eager to believe that we do live in a "small, small world." The small world problem appears to have disappeared from introductory psychology textbooks, I learned. We psychologists need to take it back.


My paper is also intended as a cautionary piece as the new efforts at replication in the computer age are making the same errors as Milgram did.  The Columbia Small World Project, for example, is confined to people on e-mail, surely a select and relatively high income group. I noticed with amusement that the Cox News Service column of January 22, 2002 emphasized that a letter had gone from Australia to Siberia in only four e-mails. This vivid anecdote is analogous to Stanley Milgram's Psychology Today piece  where the letter from the Kansas wheat farmer to the wife of the divinity student got through in only four days.  An arresting anecdote does not substitute for sampling and statistics, though it sure sticks in the mind.  I remembered Milgram's story about the divinity student for thirty years!


The Columbia  Project is also advertising in the New York Times for respondents, the same sampling error Milgram made in advertising in the newspaper rather than making any attempt at random sampling.  (He also chose blue-chip stockholders to reach a Boston stockbroker, not exactly an unbiased selection).   When my paper came out, I got an e-mail from my publisher in Seattle, saying he had responded to the Columbia Project after seeing an ad in the New York Times.  He had responded because he had an international distributor, he said, and this international distributor had  many international connections and thus might be able to get a letter through.  He told me he himself was an example of the selection bias I was writing about.


If you read the paper, you will also see that I had some empirical basis to "speculate" that Milgram's sampling methods would have been far more likely to attract people who could get their letters through.  I also had some empirical basis for the speculation that some studies might have had disappointing results and not been published.  I found several unpublished attempts at replication in the Milgram archives, which led to my "speculation' that others had tried to replicate the study and had never published their studies. At least one of these efforts was of such low quality that it shouldn't have been published.


But all this is nitpicking about an old study.  I hope Milgram is right.  I would like to believe that we are all connected to each other.  Or  is this just a "group grope," a romantic attempt to establish bonds in what we used to call at Wellesley "the cold, cruel world"?


We need new research, and I welcome it.  I have no dog in this fight.


And, thanks to Tom Blass, for saving me from making some uncalled for and unkind remarks about Milgram's feet of clay. He recalled me to my better self!!