Nearly two weeks ago, I tried a brief experiment to test Eli Pariser’s contention from his recent book, that Google’s use of 57 signals to personalize someone’s search results leads to a filter bubble.
I also promised to include a second post with more results from readers. Below are their screenshots along with those posted on Saturday.
I also included some shameless plugs for their blogs.
Here are a young female’s results from her personal computer in Idaho:
Here are that same female’s results from her laptop in Idaho, which has been “corrupted” by her parents:
Below are my results:
Here are my baby-boomer, liberal mother-in-law’s results:
With each permutation of this exercise, Eli Pariser’s claims appear to weaken. That is not to say his claims are false, just that the effect he writes about does not seem to appear strongly in this very small sample size of four.
Of course, now I want to read his book more than ever.
In terms of differences, reflectionsephemeral was the only participant who did not get a top ten search result of “War Games.” This result would bolster Pariser’s claim about political segmentation, were it not for the fact that my decidedly anti-war mother-in-law did receive “War Games” as a search result.
Another difference is that the band War did not appear as a top result for reflectionsephemeral, Vern, and me. My only guess is that reflectionsephemeral and I are not as strong consumers of music as perhaps my mother-in-law and pino are.
For some inexplicable reason, Vern, the only non-U.S.-based participant, got a link for World War II.