Google Trends: It’s Cool to Stay in (Public) School

In Socialnomics this week, I found Erik Qualman’s discussion of the use of Google search information, in the form of Google Insights, fascinating. The idea that “Google flu trends” could track and display spikes in the flu virus faster than the CDC can is amazing. (I’ll have to keep an eye on Upper Michigan to see if I really want to risk going there in two weeks; their flu activity is “moderate” right now while Wisconsin’s is “low.”

Even more interesting is that the outcome of political races is being predicted by the volume of Google searches for a particular candidate. Or that candidates might design their posters or yard signs based on whether their first or last name is more commonly searched on the internet. Incredible.

Then I wondered, Is that technology and information available to me? An average guy with no political designs and no money to subscribe to such service? Incredibly (at least to me anyway) it is available. A quick Google search found a new version of Google Insights called Google Trends. All I had to do was sign in with the Google email and password I already had.

This is basically how it works: You type in up to five search terms to compare. Google compares the number of searches for those terms with the total number of Google searches to come up with a relative volume of searches for each of your terms. The standard search shows the change in interest for your terms from 2004 to the present, but you can restrict the time frame to smaller periods according to your needs. You can also restrict the search by region, to include the whole world, a specific nation, state, or even a city. Neat, huh?

In addition to looking for the relative volume of searches for your desired terms, you can look for the number of times your terms appeared in Google News stories.

Trend information is updated daily. Hot Searches information is updated hourly.

The information is displayed in easy-to-read graphs and color-coded maps, and can be exported as a .csv file which can be opened by popular spreadsheets.

According to one of the site’s administrators, though, the information is meant to be interesting and entertaining, but not to be used for Ph.D. dissertations.

Alright. Google is offering me all this information, so what do I do with it? Should I start trying to analyze search information for candidates running for November elections? Should I try to predict results and compare my analysis to the actual outcome in November? No way! I don’t want to become panicky or depressed. At least not on that magnitude.

Instead, I tried to predict if I will still have a job in the next few years. Nothing too panic-provoking about that I hope. In light of all the concerns about collective bargaining for teachers, teacher evaluations, testing, school performance, vouchers, charter schools, online schools, and a fear of the overall decline of public education, I searched for trends relating to my job security. I entered the terms “public schools,” “private schools,” “charter schools,” “online schools,” and “homeschool.” My thought was that if searches for public schools were lower or about the same as the other types of school options, my teaching time was drawing to an end. If Google Trends are correct, public schools aren’t going away any time soon.

Google Trends chart comparing volume of searches for types of schools. 

The information was very interesting. From 2004 to the present, in the United States, public schools had an average search volume of 63, while private schools averaged 6, charter schools averaged 3, online 1, and homeschool 7. While interest in public schools fell from a relative volume of 79 in 2004 to a volume of 65 now, only interest in online school increased–from 0 in 2004 to 1 at present. This surprised me, since I thought a lot more students might now be intrigued by the idea of online education.

When the nation-wide information is broken down by state, though, in Wisconsin, homeschooling and online schools show a moderate level of search interest, while public school interest is relatively low. Broken down further, by regions of the state, the most interest in all types of schooling comes from south-eastern Wisconsin (Milwaukee schools?) and north-western Wisconsin. The only exception to that is that Madison area had the highest interest in homeschool.

Despite the growing interest in online schools and homeschool, this trending leads me to believe that public schools are still “safe” in Wisconsin. Highschoolers who try online and homeschool options ofter come back to the public schools after they find out it’s not much fun to be isolated from their peers. At any rate, we will continue to have public school jobs, but as other school options become more popular, public schools and teachers will be under greater pressure to perform–probably with fewer resources. That’s a trend I’m not happy about.

About Rob_Henseler

Rob has been teaching high school English and Language Arts for 20 years. When he's not at school, he enjoys making and listening to music, woodworking, canoeing, and hands-on traditional skills.

Posted on October 7, 2012, in Social Media, Society, Workplace. Bookmark the permalink. 9 Comments.

  1. Thanks for the link, Rob. That is a really cool site and I’ve never tried it before. I also wanted to see how it worked, so I typed in “Obama” and “Bieber”. Good news, America, the Biebs is declining rapidly in popularity and he was just recently passed by Obama.

    I don’t want to read too much into it, but any trend line that shows Bieber declining can’t be a bad thing.

    It really is remarkable that so much data is available to average folks like us. Sure, it can’t tell you if people are searching for these terms because they love it or because they hate it, but I guess if you are from the school of there’s-no-such-thing-as-bad-press this is still invaluable.

    One thing to consider, though, is that if this is the kind of information that anyone can get, what is the data the pro’s keep locked up? I recently read “The Power of Habit” by Charles Duhigg, and he relates a story in there about how the data miners at Target can predict when their female customers are pregnant, sometimes before they have even told their family. You can read the story in this NY Times article:

    • Bob,
      I loved the Target article in an amazed and somewhat frightened sort of way. Kind of makes me want to start buying products with cash only, and not letting the clerks swipe my store card. I’ve always know that the grocery store was storing my data and customizing coupons for me, but I never really pictured department stores doing that too. Maybe I’m naive, but I don’t think Fleet Farm and Menards do that. Their weekly adds are so big, they seem impossible or impractical to customize. Perhaps this is another reason to buy from the local family-owned businesses: they can’t afford to employ statisticians that the nation-wide chains can.

      I’m a homebrewer. No shame in that, but in my brew supply store, there are books about making moonshine and building stills. Might a person feel a little scared to buy such a thing with a credit card? Might a person not want that purchase to be tracked?

      On the one hand I’m a little bothered that stores pry into our lives so successfully and are so adept at manipulating our behaviors. On the other hand, I’m impressed with the genius required to read the stories of our lives so clearly and turn that information into profit.

      Sounds like an interesting ethics debate.

  2. Interesting post, Rob! I wasn’t aware that Google Trends functionality was available to the general public. It’s probably a good idea for me to stay away from trying to project November’s election results too—panicky and depressed, indeed.

    One thing that’s interesting to consider, though, is whether people tend to search more for things they’re relatively unfamiliar with. For example, I might know what I need to know about public schools—I attended them and know how they work. I would guess a lot of people search for things they’re unfamiliar with to determine their feelings about it. For a political example, take Obama and Romney. Obama has been in office for four years; people know him. People may not have been as familiar with Romney, on the other hand, resulting in more internet searches. The same could hold true for public schools versus other education options. Maybe fewer people search for those things because they are familiar.

    • You make a good point about familiarity and how it would influence searches. I like how Qualman described the intricacies of analyzing the trend data with some of that familiarity in mind. My analysis was purely from the perspective of an amateur. As one of the representatives of Google Trends said, the information is for interest and entertainment only, not to base a Ph.D. dissertation on.

  3. Rob, thanks for taking the extra step and exploring on Google Trends. I was wondering about that too. We just have to keep in mind that statistics can be interpreted not just in one way.

  4. The Google Trends chart you link to in your post isn’t opening for me, but I really like the set up you give to how this works as well as the explanation you provide about your sample search.

    And as for Bob’s point about Bieber, look at how Twitter changed their algorithm to prevent him from dominating the conversation:

    • Sorry about the Google trends link. I spent a lot of time trying to get the chart directly into the blog post, but that didn’t work. As it turns out, this didn’t either. It really is a neat site, though.

  5. Also, do we know why your post only appears as a “Read this entry” link on the home page rather than the entire text showing?

  6. Is it showing as “Read this entry” because it was a bit long? That was my guess, since I did nothing different than the other posts (that I’m aware of) except writing a bit more.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.