Meta-Analytic Learning Disability Research is a lot like Spotify
By: Mia Daucourt
Arguably, one of the best features of Spotify, is the fact that it chooses music for you. It literally anticipates your needs. With so many competing possibilities in the vast realm of music, it’s refreshing when Spotify uses the information it has about you to predict your needs based on your music listening history and preferences. That way, after you’ve gotten to the end of playlist, whether it is of your own making or one of Spotify’s many radio stations, Spotify will not just stop playing music; instead, it will continue to play songs that it predicts you will enjoy. Spotify’s predictions are not always on the nose, but the more information you have provided about your preferences, like indicating which songs you like with a heart that is much like Instagram’s double tap, the more accurate its predictions tend to be. This idea of using what you know about someone to anticipate their needs is quite a luxury in terms of music streaming, but for learning disabilities research and treatment it is absolutely pivotal. Having and using what you know about someone’s learning needs to anticipate other learning-related struggles that may arise is important in a realm where children are 4-5 times more likely than children with no learning-related disorders to have an additional learning-related disability when they already have one.
Meta-analysis is a research technique that aligns with Spotify’s information gathering approach to prediction, making it especially well-suited and powerful for learning disability research focused on predicting learning-related needs. Specifically, meta-analyses require that a researcher gather as much empirical evidence as possible on a chosen topic area, in order to figure out, based on lots of data, how two realms are related. In Spotify terms it would be like looking across all of Spotify’s members to see how closely related liking one song is to liking another song, based on how many people listened and/or indicated a preference for both. A recent meta-analysis by Daucourt and colleagues (2019) used all the twin studies conducted on reading and math and reading and attention deficit hyperactivity disorder (ADHD) to figure out how many common genetic and environmental risk factors could be found between reading and math and reading and attention. In other words, how likely were people with reading problems to also have problems with math versus problems with attention? This information could then be used to create screening questionnaires that help researchers and educators anticipate children’s non-reading learning needs based on how they do in reading, before they fall too far behind. Daucourt and colleagues found that reading and math had more common influences than reading and attention, meaning that screeners that looked for reading problems should be including questions meant to identify math difficulties.
In addition to looking at the links between songs, Spotify is also likely to gather more macro-level information on user characteristics to help inform their music predictions. For example, listening patterns may show that certain countries are more likely to prefer certain music genres or artists. Take David Hasselhoff’s career as a popstar in Germany, as an example. Based on a country-specific phenomenon of “the Hoff” making it big in Germany, a listener in Germany is probably more likely to enjoy a David Hasselhoff song showing up in their Spotify queue than a listener in the U.S. would.
In the same vein, Daucourt and colleagues’ meta-analysis also looked at how the genetic and environmental links between reading and math and reading and ADHD symptoms differed based on the sample, method, and measurement characteristics that varied among the research studies they gathered. Much like Spotify would likely find between Germany and the U.S., Daucourt and colleagues found significant differences between countries and twin projects that study reading and math problems and reading and ADHD. This revealed that when we investigate the co-occurrence of learning-related abilities it is important to look beyond a single sample, which may have special characteristics that do not generalize to a wider population. Meta-analyses that take differences between studies and samples into account are especially important in twin-related research because the twin samples currently used in learning disability research seem to be quite different from one another. Accordingly, Daucourt and colleagues’ work will hopefully spark a trend to be more like Spotify in our efforts to understand learning disabilities, how they co-occur, and which characteristics influence how they are related.
Citation: Daucourt, M. C., Erbeli, F., Little, C. W., Haughbrook, R. & Hart, S. A. (2019): A meta-analytical review of the genetic and environmental correlations between reading and Attention-Deficit/Hyperactivity Disorder symptoms and reading and math, Scientific Studies of Reading
, DOI: 10.1080/10888438.2019.1631827