top of page
  • Writer's pictureAshley Waldron, PhD

Peanut Butter and Reproducibility

Perhaps you’ve heard the quote from Isaac Newton “If I have seen further it is by standing on the shoulders of Giants.” This sentiment has come to represent the ideal that researchers strive to meet. We rely on the observations, conclusions, and hypotheses made by those who came before us to help us develop our own ideas. But what happens when it turns out that those “Giants” were not solid beings, but structures of straw, waiting to be blown down by a gust of skepticism? All that time spent developing your hypotheses and conclusions could have been wasted.


How do we ensure that the shoulders we stand on are stable? And perhaps more importantly, how do we ensure that our work is forming a solid base for those who come after us? Short answer - we can make sure that our research is reproducible. Being able to reproduce the results of a previous study makes it much easier to have confidence in the data and ultimately the conclusions.


Reproducibility encompasses a number of practices and aspects of the research process. One aspect is to provide the necessary methodological details so that someone else could run the same experiments or analyses and produce the same results. Critical to achieving this goal is keeping good notes for yourself and then communicating those notes effectively to your peers.


It takes time to figure out the level of detail you need to include in, for example, a methods section of a report. One way to get started, even if you’re not in a lab yet, is to imagine making a sandwich. (Or try it for real! This is a pretty common exercise used in classrooms to explore communication and instructional writing and it could easily be applied to science courses).


Tastes like reproducibility


Imagine a very persnickety guest has arrived at your house and is hungry. Read the following passages and consider which one you would prefer if that guest requested the “Perfect” peanut butter and jelly sandwich and provided you with a recipe from their personal chef that makes it just right.


Questions to consider:

  • Which description would you find most useful for preparing the “perfect” sandwich?

  • What details were included in the option you chose?

  • How do these descriptions compare to a methods section you may have read or written in a report? Recipe 1: Prepare a peanut butter and jelly sandwich as usual. Recipe 2: Apply chunky peanut butter and raspberry jam to two slices of bread. Put the bread slices together and prepare for serving. Recipe 3: Place two slices of Oatmeal Bread side by side on a plate. Spread to within ¼ inch of the crust Teddy’s Extra Chunky Peanut Butter on one slice of the bread. On the other slice, spread Bonne Maman Four Fruits Spread to within ¼ inch of the crust. Press the two slices together so that the peanut butter face is in contact with the jam face and the crusts are aligned. Cut the resulting sandwich at a diagonal to create two triangular pieces before serving.

What does this have to do with science?


Hopefully it is clear that you would have a better chance of creating the “perfect” sandwich using Recipe 3 than either of the first two. It may sound silly to think about making a sandwich, but we encounter similar types of scenarios in research all the time. Sure, the methods section of that research paper you just read may not have sounded like a recipe, but it serves a similar purpose. A recipe tells you what ingredients to use and how to use them to create a particular dish; methods tell you what materials to use and how to use them to produce a particular result. And just like a recipe there will be some details that you leave out of an experimental method, such as the brand of pipettes you used or what you were wearing that day.


The difference in tone and style of a methods section versus a recipe or a protocol is another story for another day. Today, I just hope you take away a better idea of the type of details you should record, share, and look for in others research and why it matters. Issues with reproducibility can waste time and resources and undermine the credibility of entire fields of research. If making a simple peanut butter and jelly sandwich can be a source of confusion, imagine a complex experimental protocol. Having access to detailed methods helps avoid confusion and promotes stronger discoveries.



Resources



Comentários


bottom of page