A Blog by Jonathan Low

 

Jul 5, 2015

Days of Our Digital Lives: Learning Why We Search For What We Search For When We Search For It

The line between our digital lives and whatever that other, offline life might be is rapidly disappearing. So the patterns we establish by what we search for when - and why - provide increasingly important insights into how we live and what motivates us. JL

Seth Stephens-Davidovitz comments in the New York Times:

One of the most important powers of big data is simple. Just as the microscope allowed us to look at smaller and smaller parts of physical objects, big data allows us to zoom in on our minds.
WE search for doctors at 8:20 a.m., recipes at 4:52 p.m. and enlightenment at 3:16 a.m. So far, the Google data that analysts have worked with has been daily, weekly or annual, but Google recently released the same anonymous, aggregate data down to the minute. I looked at this data for New York State for four weekdays, a small sample, certainly, yet the patterns were very sharp.
Google search rates for “unblocked games” — those that are not blocked by school administrators — peak at 8:04 a.m., stay high through 1:30 p.m., and drop precipitously afterward. (Score one point against cellphones in schools.)
Search rates for “weather,” “prayer” and “news” peak before 5:30 a.m., evidence that most people wake up far earlier than I do. Search rates for “suicide” peak at 12:36 a.m. and are at the lowest levels around 9 a.m., evidence that most people are far less miserable in the morning than I am.
The data shows that the hours between 2 and 4 a.m. are prime time for big questions: What is the meaning of consciousness? Does free will exist? Is there life on other planets? The popularity of these questions late at night may be a result, in part, of cannabis use. Search rates for “how to roll a joint” peak between 1 and 2 a.m.
Late night is also a time rich with anxiety. It is the time with the highest rate of searches that include the word “symptoms,” a worrisome category that captures searches for heart attacks, colon cancer, H.I.V., A.L.S., strokes, brain tumors and many other conditions.
One very important weakness in this data: Google releases only search rates, not the total number of searches. If a search rate for a word is highest at 3 a.m., it means that of all the searches at that time, an unusually large percentage of them include that word. Another important weakness of the data: Our searches most definitely do not capture every activity. If you have a regular yoga appointment at noon, you may not need to make any searches about it.






“Porn,” which is three times more popular among men than women, is most popular from midnight to 2 a.m. “Literotica,” which appeals primarily to women, has a substantial early-morning presence as well (it is most popular from 3 to 5 a.m.). Search rates for “vibrators” are highest at about 4 a.m. Search rates related to condoms, such as “how to put on a condom,” peak at 10:28 p.m.
The question “why is my poop green?” is asked most between 5 and 6 a.m. and between 6 and 7 p.m.






There is some evidence that we get less sharp as the day progresses. Between 2 and 3 a.m., search rates for “forgot password” are 60 percent higher than average. They are lowest around 9 a.m. Between 2 and 3 a.m., we are more than twice as likely to misspell “facebook” as “facbook” and nearly twice as likely to misspell “weather” as “wether.”
While my main analysis used New York data, I did briefly test search patterns in other areas. Mostly, people in different parts of the United States and the world are very similar in how we search. For example, every part of the world I looked at consumes more news first thing in the morning than at any other time of the day.
One interesting cultural difference I found is in what we do during lunch. Which searches spike around 12:30 on weekdays? In New York and most places in the United States, there does not seem to be a consistent lunchtime activity, but in other countries there are clear patterns. In Britain, people catch up on the news. In Japan, there is a noticeable rise in travel planning. In Belgium, it’s anything shopping related.

None of these findings shake the foundations of our understanding of the human psyche, but we shouldn’t underestimate the value of the data Google releases either. Minute-by-minute analysis like this is a relatively new treat and it can be very useful to scholars: The economists David Card and Gordon B. Dahl analyzed crime data minute-by-minute and found that domestic violence complaints rise immediately after a city’s football team loses a game it was expected to win.
One of the most important powers of big data is simple. Just as the microscope allowed us to look at smaller and smaller parts of physical objects, big data allows us to zoom in on our minds.

0 comments:

Post a Comment