1
$\begingroup$

I want to compare the popularity of several fish types: The way I'm doing it is by using a reference keyword (in this case "fish") and comparing it to whatever fish I'm interested in. I enter both keywords (fish and tuna or fish and salmon etc) in Google trends and it gives me an average of both keywords. Then I simply divide these averages and get an OR. I keep repeating this process for each fish type. Can I do that? And would that be called an Odds ratio?

I'm hoping that this way I can get a feel of how popular each fish type is.

$\endgroup$
  • 2
    $\begingroup$ Why do you feel the need to compute an odds ratio to compare averages? $\endgroup$ – StatsStudent Jan 26 '19 at 17:06
1
$\begingroup$

The y-axis values returned by Google Trends are not odds -- it's not clear what exactly they are, to be honest -- so their ratios are not odds ratios.

$\endgroup$
  • $\begingroup$ No, but they do compare two keywords when I enter them. I'm thinking that if I use the word fish as a reference then I can calculate how popular a given keyword is compared to the word fish. Can I do that? Maybe it's not called odds ratio then? $\endgroup$ – Blits Jan 26 '19 at 17:01
  • 2
    $\begingroup$ Right, I see no reason why you can't compare the popularity of these terms using the metrics they give, I just don't think "odds ratio" is an appropriate name for those ratios. $\endgroup$ – Jake Westfall Jan 26 '19 at 17:03
  • $\begingroup$ Moreover, whatever Google trends returns, it is standardized. $\endgroup$ – Tim Jan 26 '19 at 17:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.