01
About Dustin
Founder of Simplistics
Dustin Fife was…ah, who’m I kidding. We all know I am writing this. Hi, I’m Dustin. Welcome to simplistics!
I received my PhD in Quantitative Psychology in 2013 from the University of Oklahoma, which makes me a Quantitative Psychologists. And what is that? It’s essentially a statistician in psychologist’s clothing. Or, it’s a statistician with training in psychological measurement, experimental design, and non-experimental methods.
I did my dissertation on missing data, but since migrated my research interests. Now I primarily specialize in developing tools for data visualization, with a secondary research interest in statistical pedagogy. I’ve published in many big-name flagship journals that nobody but scientists have heard of, including Psychological Methods, American Psychologist, and Perspectives on Psychological Science. I’ve also written a textbook that is entirely too hilarious to be educational called The Order of the Statistical Jedi: Rights, Rituals, and Responsibilities.
I currently work as Rowan University in Glassboro, NJ, where I direct the research methods/statistics training of doctoral level clinical psychology students. I also have a YouTube channel called Quant Psych.
02What is Simplistics?
Simplistics is short for “simplified statistics.” But, maybe that was obvious.
Here’s the thing: scientists have been trained using the exact same statistical methodology since the 1970s. This methodology is primarily concerned with teaching students what “tests” to use. Is it a t-test? An ANOVA? A regression? Chi square? Once they know the test they shoud use, they’re trained to hunt for the p-value so they can conclude if their results are “statistically significant.”
And that’s it.
That’s an awful way of analyzing data. This entire curriculum was developed before computers! So, our entire curriculum that we currently use was developed before we even knew what a calculator was?
Do you see the problem?
One-sample t-tests, ANOVAs, independent t-tests–they’re all mathematical shortcuts developed to make hand calculations easier. And they’re all the same basic procedure called a linear model. Everyone knew in the 1970s that t-tests/ANOVAs/factorial ANOVAs/ANCOVAs/etc. were all special cases of the linear model. But once computers made these hand calculations unnecessary, the habit stuck and we still treat these procedures as separate and distinct when we don’t have to!
The simplistics approach recognizes that these distinctions complicated something that doesn’t need to be complicated. Once we realize it’s all the linear model, analyzing data becomes immensely simplified.
03
Why does it matter?
We need to teach statistics differently. “But why?” you might ask. “Yeah, it may be a bit harder to do it the old way, but who cares? It builds character!”
Well, so does getting punched in the face.
The problem is that the consequences of the old way have been severe. Entire bodies of literature (e.g., medicine and psychology) have been discarded because people realized our reliance on the old way of doing statistics is extremely problematic. You see, the old way, with its obsessive focus on p-values, encouraged researchers to game the system. Afterall, if the only statistic you rely on is p-values, then there’s only one statistic you need to satisfy.
My approach reinforces a much more holistic way of looking at data. We might look at model plots, diagnostic plots, means, mean differences, slopes, AICs, BICs,
When you use my approach to analyzing data, you protect yourself from being deceived. You protect yourself from buying into your beloved hypothesis at the expense of truth. Equally important, you guard against missing important and interesting information that was waiting to be discovered.
Your data is like a rich birthday cake. The old way of doing things is like pulling a single sprinkle off that cake. So much richness remains! And with simplistics, I’m going to teach you how to ingest the whole thing (without having to suffer a diabetic coma!).
