Inappropriate Uses of Google Trends
medium.com
Inappropriate Uses of Google Trends
1–10 of 113 posts
Re: Inappropriate Uses of Google Trends
#2In 2014, Facebook made a famous post (https://www.facebook.com/notes/mike-develin/debunking-prince...) mocking researchers for using Google Trends data as evidence it would die.
I made a comment along the lines of this a year ago (https://news.ycombinator.com/item?id=9828295) when people cited Google Trends as statistical evidence people were looking for a Reddit alternative in light of the controversy back then.
Google Trends is fun for snarky Tweets, but I'm still seeing it used to prove points in startup blogs, which is a problem.
Re: Inappropriate Uses of Google Trends
#3I'm glad this was written and has more historical examples than just Brexit.
I could imagine a good portion of the questions could be originating from students. If a large political or economic event happens, it's not unlikely for classrooms teachers to task their students to understand what's happening. Search is the obvious first step for that.
Even more broadly, and in line with the article, extrapolating from a single coarse data point is a Bad Idea™.
Re: Inappropriate Uses of Google Trends
#4In 2014, Facebook made a famous post ( https://www.facebook.com/notes/mike-develin/debunking-prince... ) mocking researchers for using Google Trends data as evidence it would die. I made a comment along the lines of this a year ago ( https://news.ycombinator.com/item?id=9828295 ) when people cited Google Trends as statistical evidence people were looking for a Reddit alternative in light of the controversy back then.…
It's easy for Facebook to say, here's our actual engagement numbers because we have (nearly) perfect tracking! But from an outsider's perspective, say a journalist or a hedge fund bro, it's difficult to perform any truly rigorous analysis involving these unobservable behaviors. So people come up with proxies - Google Trends, polls, or if you're truly adventurous, credit card spend information. [1]
Anyways, using products like Google Trends is worth a shot if only because there's few substitutes to imperfect information.
[1]: http://www.reuters.com/article/us-sec-capitalone-insidertrad...
Re: Inappropriate Uses of Google Trends
#5I'm glad this was written and has more historical examples than just Brexit. I could imagine a good portion of the questions could be originating from students. If a large political or economic event happens, it's not unlikely for classrooms teachers to task their students to understand what's happening. Search is the obvious first step for that. Even more broadly, and in line with the article, extrapolating from a s…
If you start with the assumption that your enemies are stupid, you won't have much trouble finding the evidence.
Re: Inappropriate Uses of Google Trends
#6Good piece. The conclusion on the Brexit article was ridiculous.
Re: Inappropriate Uses of Google Trends
#7You mean to tell me clickbait media is biased and manipulative?
I'm shocked.
Re: Inappropriate Uses of Google Trends
#8The Washington Post reporters knew that Google Trends data was not accurate and misleading. They went with it because: it is a liberal site, wanted to make people angry against the Brexit voters. Also Jeff Bezos(the owner) has a beef with Trump and publicly supports liberal media outlets like Gawker.
tldr: Nothing unexpected happened, they knew what they were doing.
Re: Inappropriate Uses of Google Trends
#9Similar-ish to the cancelled 'Google Flu Trends' project, which tried to predict flu outbreaks based on search trends. A paper was initially published, containing fantastic results, but since then the results have not been great:
http://www.npr.org/templates/transcript/transcript.php?story...
I like the Google quote at the bottom of the article - '...part of the issue was that they sometimes get a flood of searches simply when there's a lot of flu in the news'.
Who would have thought of that, eh? It apparently took eight years to discover!
Re: Inappropriate Uses of Google Trends
#10In 2014, Facebook made a famous post ( https://www.facebook.com/notes/mike-develin/debunking-prince... ) mocking researchers for using Google Trends data as evidence it would die. I made a comment along the lines of this a year ago ( https://news.ycombinator.com/item?id=9828295 ) when people cited Google Trends as statistical evidence people were looking for a Reddit alternative in light of the controversy back then.…
It's easy for Facebook to say, here's our actual engagement numbers because we have (nearly) perfect tracking! But from an outsider's perspective, say a journalist or a hedge fund bro, it's difficult to perform any truly rigorous analysis involving these unobservable behaviors. So people come up with proxies - Google Trends, polls, or if you're truly adventurous, credit card spend information. [1] Anyways, using prod…
Printing misleading charts and presenting as fact is often worse than printing nothing at all.