The 2018 Dictionary.com word of the year was “misinformation," which — in retrospect — is unsurprising, given the ever-escalating volume of false claims propagated both online and in real life. The COVID-19 pandemic brought with it new false claims, some about the fabricated dangers of vaccines, and others about unproven treatments. Underlying all of these false claims was the diminishing value of evidence, and a deterioration of evidence evaluation skills. As teachers, we cannot underestimate the importance of teaching our students the value of evidence-based thinking, nor can we start teaching these skills too early.
This fact is true not just for “hard sciences” but for all disciplines. English teachers ask students to use textual evidence to support claims about works that they read. Comparative Politics students examine data like political participation, electoral processes, and political pluralism to evaluate the degree to which a country can be considered democratic. In Biology, students are taught to systematically observe living organisms and record their observations to answer questions about how these organisms live and relate to their habitats.
But in Data Science, we not only put evidence on a pedestal; we also shine a bright light on it. We teach our students the following skills, among others:
- to critically examine the source of their data. Where does the data come from? How was the data collected? How was the sampling conducted? Is the sample representative of the population? Answering these questions helps to evaluate the credibility of the source.
- to effectively visualize their data. What type of graph (histogram, bar chart, scatterplot, etc.) would work best to communicate insights about evidence in an understandable and interpretable way? What common practices of misrepresenting data visually should we be aware of?
- to interrogate measurement validity. Is the variable measuring what it claims to measure? Is there a better measurement available?
- to examine variables one at a time as well as collectively. How can we understand data characteristics, patterns, and anomalies? What does a relationship between two or more variables look like?
- to consider the ethics of data. How can we think of evidence of biases or discrimination? Are there privacy concerns in handling particular data? Are any groups systematically excluded from data sets? What are the implications of such exclusions?
- to model data. How do we make generalizations from data sets? How can we use these generalizations for further predictions and forecasts?
We elevate empirical evidence to inspire our students to be better problem-solvers and decision-makers, to produce high quality research, to continually work on reducing their biases and subjectivity, and to continue to advocate for a transparent and just society for all. Perhaps we can hope for a future in which the word of the year is “evidence.”