The Need For Flexible and Nuanced Recommendations
By Megan Sumeracki
Last week, Emily Oster wrote a ParentData newsletter titled What to do when what’s best is impossible: How to embrace the “second best” option. The main idea is that in parenting, and in life, there are recommendations based on some perfect ideal and this is the first best option. But when that option just doesn’t work, or is impossible to achieve, we (parents, or really anyone who isn’t spending a ton of time reading every research paper on a given topic) don’t really know what second best is.
[By the way, if you don’t follow Emily Oster’s work, I highly recommend her free ParentData newsletter (and for a small fee, if you can afford it, you can access her massive archive). I personally follow her, and search her archive regularly while trying to solve my own parenting problems with my daughter. In my opinion, she does a wonderful job of presenting data (and identifying when we don’t have good data) in a way that is clear and concise without eliminating the nuance.]
Photo by Megan Sumeracki
Here is an example she gives in the newsletter: baby sleep location. The first best is that babies sleep alone in a crib, on their back with no loose blankets or toys, in their parents’ room, for at least the first six months of life. This is what the AAP recommends and is considered the gold standard in the United States.
However, what should a parent do when this simply isn’t possible or feasible? Emily Oster points out that if you’re in a situation where the best isn’t working for you, it can be difficult to figure out what the second best really is. For example, co-sleeping on a couch is much more dangerous than co-sleeping on an adult-sized bed with no blankets or pillows. Here’s one I’ve also wondered about: Is there really a safety difference between putting your baby to sleep in their own room with a video monitor and putting them to sleep in your room? Surely both of these are safer than sleeping together on the couch… right? This information can be difficult to find, as the recommendations tend to stick very closely to the gold standard, and not provide any other “second best” options or nuance.
Emily Oster’s overarching message in this newsletter really resonated with me (and not just because I love Emily Oster). The newsletter calls for more information about second-best options so that parents can make informed decisions, even when the first best does not work for them. It calls for recognizing reality and including more nuance within the advice that is given. I love this because in many ways this is what we are trying to do with The Learning Scientists.
If you’re a frequent reader of our blog, you know that we are big fans of describing flexible guiding principles rather than giving more specific prescriptions. Why? Because the specific prescriptions would not work for everybody and would not work in every situation! The science of learning is nuanced, and so the recommendations also need to be nuanced. Further, we feel strongly that it is important for students, teachers, and other educators applying the science of learning to understand the underlying principle driving the effective learning strategy. When the underlying principle is understood, then adjustments can be made to how the learning strategies are used in practice to best fit the given situation while remaining effective.
Take an example: retrieval practice. A very concise and specific recommendation, based on centuries of data on retrieval practice, could be to test more in school. There was a time when most research on retrieval was labeled “the testing effect” and recommendations focused on testing as a way to promote learning. This recommendation is not bad, per se. Saying “test more” is simple and clear, and when done well would likely lead to effective and efficient learning. But if all we say about retrieval practice is that students should “take more tests”, then when this doesn’t happen retrieval practice will be pushed aside entirely. Testing has a lot of negative associations, and when we say “testing” a lot of students, teachers, and other educators think of standardized testing. By and large most do not want more standardized testing. Further, high-stakes tests might not produce as much learning as low-stakes or no-stakes retrieval opportunities due to test anxiety. Without even knowing it, people end up throwing the baby out with the bathwater, so to speak.
Instead of going for the simple yet stripped-down recommendation, we feel strongly that those engaged in science communication should present the nuance. We think it is much better to describe a flexible guiding principle, bringing information to mind from memory (retrieval practice), along with a number of concrete examples (like these blog posts: retrieval mapping, quizzing format, reflective class feedback, and optimizing learning with retrieval practice). The nuance here is that testing is not the only way to implement retrieval practice. There are plenty of ways that retrieval can be infused into learning activities that don’t look at all like tests. There are also ways to reduce the pressure during tests or quizzes, and this type of low-stakes or no-stakes practice can help alleviate test anxiety (I talk about why it is a good idea to still include some tests/quizzes here).
It takes a bit more time, and it means that the advice that we give is sometimes not the most prescriptive that it could be. But, by explaining the nuance and describing flexible guiding principles, students, teachers, and other educators can make informed decisions about what is best for them/their students.
