Skip to main content

Posts

Showing posts from July, 2020

CNN's "Science by press release"

One of my big pedagogy concerns, as a psychologist who teaches psychology majors, is this: Are we explicitly drawing the links between psychological science and ALL of the other sciences, and the fact that many of the lessons they learn in their psychology classes apply to other sciences?  I think this is an issue in statistics. I always emphasize that I do not simply teach statistics for psychologists: I am teaching them statistics, full stop. I think we also have to emphasize to our majors that the psychology research process is, in many ways, just the broad research process use in science. As such, our lessons aren't just teaching them major-specific content, but we are teaching them information that leaves them better prepared to interpret scientific research they encounter.  This includes a potential ugly part of the research process: Bad science reporting via over-hyped research press releases.  As such, I present this great piece from CNN, "Science by press release...

The only data set you'll ever need: Nathan's Hot Dog eating contests, 1908-2019

Physiologist Dr. James Smoliga published an article entitled " Modeling the maximal active consumption rate and its plasticity in humans - perspective storm hot dog eating competitions. " While hot dog eating competitions may not seem germane to Serious Academic Discourse, the idea of stomach/gut plasticity certainly is. Spoiler alert: According to the models, the maximum capacity of a stretched-out human stomach is 84 hot dogs. And buns.  Honestly, all of the GIFs of humans eating hotdogs were nasty, so enjoy this cutie. However, my blog post is about something other than the researcher's findings as much as it is about  Nathan's Hotdog Eating Contest spreadsheet that Smoliga created while performing his research. A spreadsheet packed with 16 variables and 430 hot dog eating participants your students can analyze in Stats class. I'm surprised that Nathan's didn't have its own database. Here is a description of how they generated the database.  Independent...

NPR's The Pandemic Is Pushing Scientists To Rethink How They Read Research Papers

This NPR story by Richard Harris describes science's struggle to keep up with the massive amount of COVID-19 research, much of which is coming out of China. How does science, and society, judge the quality of these papers?   How to use in class: 1.  How do scientists assess the quality of research? By reading pre-registered reports and pre-prints : The report explains pre-prints and pre-registration! The good: The research gets out faster. Reviewers can compare pre-planned analysis to the actual analysis. The bad: The media gets too excited about pre-prints. The report describes the totally overwhelming number of pre-prints for COVID-19 related research: One of the scientists interviewed in the piece describes how he used pre-registered information to assess a COVID-19 research paper: 2. How do scientists assess the quality of an article: By the author and their academic affiliation. The report describes the bias that may exist when we lean on author/affiliation heuristics i...

In-house restaurant dining is related to increases in COVID-19 cases: Illustrates correlation, regression, and good science reporting

Niv Elis, writing for The Hill, summarized a report created by JP Morgan analyst Jesse Edgerton. The report found a link between in-restaurant spending from three weeks ago and increases in new cases of COVID-19 in different states now. Data for the analysis came from 1) J.P. Morgan/Chase in-restaurant (not online/takeout) credit card purchases and 2) infection data from Johns Hopkins.  How to use in class: 1. Correlation/regression: This graph, which summarizes the main findings from the report, may not include my beloved APA axis labels, but it does include an R2 and is a good example of a scatterplot.  ALSO: The author of The Hill piece was careful to include this information from the study's author, which clarifies that correlation doesn't necessarily equal causation. 2) Creativity in data analysis: Often, in intro psych stats, we use examples rooted in traditional social science research. We should use such an example. But we MUST also use examples that demonstrate how d...