Google’s new personalised search: reinforcing bias

8 December, 2009
By Raphael Fraser

I’ve read this morning about Google’s roll-out of personalised (or personalized, for you North Americans) search, using either your saved web history, if you have a google account, or a cookie saved on your computer, if you don’t. I have real concerns about this, that don’t seem to be echoed much around the net – at least as far as I find with … a Google search ;)

Generally people are discussing it as a choice between convenience and privacy. Google certainly bang on about the convenience and relevance side of things:

The only concern raised on most of the websites I’ve found is simply privacy: Oh Noes!!1! Google saves my searches!, or Oh Noes!!1! Google puts a cookie on my computer! Frankly, I see that as close to a non-issue compared with the major potential effect: skewing and biasing the information obtained by a searcher. The examples Google give: someone searching for recipes and using epicurious.com a lot, so epicurious ranking higher in their personalised results; and two searches for “sox” one meaning baseball and one something else I didn’t catch as I was looking at another window ;) are minor, in my view. – Though if you’re a new website/company you’d not think the former was minor, I guess – which is a point that some web commenters make. That’s the convenience. Frankly, I think they’re specious examples. If the guy likes epicurious so much he ought to have it as a bookmark (or just take the two seconds to type it in the address bar). Seriously, who searches for something they already know?

More seriously, take the science/medicine/skepticism arena. If I search for autism and vaccines, for example, my personalised results are going to be very very different from Jenny McCarthy’s or Meryl Dorey’s. If I search for something to do with depression or antidepressants, my personalised results are going to be a world away from someone (a) not a psychiatrist, and (b) who might have a tendency to like vitamins and herbs and yoga etc (for example). The point is, if you’re searching for information, then the broad range of available information should be presented, not simply information (or misinformation) that confirms your personal bias.

Just as it is not Google’s role to act as scientific/skeptical watchdog and filter out any pseudoscience from their results, I submit it is not their role to in effect hide from view information that runs counter to a person’s own bias. In an age defined by information, this is a crucial issue.

I also don’t know what to think about this: I followed Google’s instructions to turn off personalisation, and found fail. :| As I have a Google account, I was instructed to go to my account, click edit next to ‘my products’ and ‘delete web history’. It doesn’t appear to be there. Now I don’t know whether that means they don’t have my web history, or whether I just can’t delete it :|

Popularity: 2% [?]

FacebookStumbleUponTechnorati FavoritesTumblrPosterousShare

Tags: , , , , , , , , ,

One Response to Google’s new personalised search: reinforcing bias

  1. [...] on destroying not search (duh) but the validity of your search results. First they came out with personalised search. This ridiculous idea means that the results you get when you search are filtered, based on the [...]

Leave a Reply

Your email address will not be published. Required fields are marked *

*

Say NO to the National School Chaplaincy Program

Tag sphere

Search the site:

Statistical data collected by Statpress SEOlution (blogcraft).