Research Projects


The Evolution Lab's research main research topics are about status/appearance; the mating market (primarily using Twitplat); and the menstrual cycle (using daily diaries).

Please find any of Dr Blake's active projects published  on the Open Science Framework (OSF).

Global Map


For researchers interested in big data and social media, our TwitPlat database can provide access to billions of data-points from around the world. At The Evolution Lab we have conceived, developed and are now maintaining a database of 6 billion geolocated Twitter posts (spanning 8 years). What first sets TwitPlat apart is the sheer size of data availability. Adding to this strength, all posts are geolocated to very fine granularity using an algorithm developed with help from a team of computer scientists. This Locatarithm algorithm geolocates posts with unique specificity, locating:

  • Every city in the world with over 100,000 inhabitants.
  • Every country in the world.
  • Using six languages.
  • Every United States city with over 5,000 inhabitants.
  • Every Australian suburb, city and region.

No other algorithm known provides such fine-scale variation in location, and it is this variation that provides us with the unique ability to examine socio-structural correlates of behaviour across time and place. If you are interested in using the TwitPlat database, please get in touch with us!

Female Hormonal Cycle

Menstrual Cycle Studies

Are you interested in understanding how female cognition and behaviour is influenced by fertility? Although this is a fruitful area of emerging research, measuring fertility and menstrual cycle phase can be indirect, imprecise and ambiguous.

To resolve this problem and help other researchers, we have developed a data-driven method for characterising women’s fertile phase. Our research compared frequently used counting methods (including several approaches to forwards- and backwards-counting) against luteinizing hormone tests. While developing these guidelines for best practice, we found:

  • No counting method predicted ovulation with more than 30% accuracy.
  • Less than 40% of the fertile days were predicted using counting methods.
  • Combining counting methods with a luteinizing hormone test (a relatively inexpensive urine test) can improve fertility accuracy to 95%.

If you are interested in conducting menstrual phase, hormonal contraceptive or fertility studies, please don’t hesitate to get in touch. You can also find more information about Standardized Protocols for Characterizing Women’s Fertility through our Open Source Framework webpage.

Image from The Menstrual Cycle at Hello Clue.