Yesterday, Google Labs announced Correlate, a new tool takes Google Trends and throws it in reverse. With Correlate, researchers instead start with a trend and correlate that data to search terms.
While the Internet giant already has tools that access search data, namely Google Trends, existing tools aren't designed to help researchers use search data to estimate real-world activities. In short, until now, your trend was only as useful as your search term. But what if you could start with a trend and see the search results that correspond to that trend? Correlate fills the gap by allowing users to upload an existing data series (trend) and see a list of search terms that relate to the trend.
In the blog announcement, Google uses the flu as an example. They uploaded several years of flu activity data from the U.S. CDC and found that people search for related terms ("cold" and "flu") in patterns that correlate to actual flu rates. Searching for a specific term (Google uses "ribosome") reveals others terms that attract corresponding activity (in this case another biological term, "mitochondria," follows a similar pattern). While this doesn't explain why the two terms follow the same pattern—correlation isn't the same as causation—it provides researchers with new avenues for research. And because Google collects data from millions of individual, anonymous searches, Correlate allows researchers tap a huge sample.
For non-researchers, though, Correlate can be fun to use. Try out the Search by Drawing tool, through which you can draw a trend line and reveal related search terms. And, if you're looking for a more animated take on Correlate, check out Google's Correlate Comic Book.
