Jamie Hale

Jamie Hale

Wednesday, January 8, 2020

Research Statistics: Frequentist & Bayesian


Learning about stats will help you think in terms of probabilities, and allow you to gain a better understanding of research data. Discussions on research stats generally involve two categories: Frequentist and Bayesian. Frequentist methods refers to quantities that are hypothetical frequencies of data distribution patterns under an assumed statistical model. These hypothetical frequencies that are predicted are called frequency probabilities. These probabilities are not synonymous with hypothesis probabilities. Bayesian statistics are also concerned with probability and present mathematical models of data. The formula use is different with Bayesian vs. Frequentist models; a key difference is how probability is conceptualized. My knowledge is in Frequentist stats; I don't have the knowledge to talk about Bayesian models, so discussions in this book, regarding stats, will be focused on Frequentist stats.

To learn more about Bayesian vs. Frequentist refer to:
Frequentism vs. Bayesianism: Jake VanderPlas- video
https://www.youtube.com/watch?v=KhAUfqhLakw

All About The Bayes: Kristin Lennox- video
https://www.youtube.com/watch?v=eDMGDhyDxuY&t=3041s

Statistical Modeling, Causal Inference and Social Science
https://statmodeling.stat.columbia.edu/

Myths about statistics https://www.statisticsdonewrong.com/

Most scientific and technical journals contain some form of statistics; that is, if the research is quantitative. Without an understanding of statistics, the statistical information contained in the journal will be meaningless. An understanding of basic statistics will provide you with the fundamental skills necessary to read and evaluate most results sections. The ability to extract meaning from journal articles, and the ability to evaluate research from a statistical perspective are basic skills that will increase your knowledge and understanding of the article of interest.
Gaining knowledge in the area of statistics will help you become a better-informed consumer. If you understand basic statistical concepts, you will be in a better position to evaluate the information you have been given.

People like assertions that reflect certainty. Statistical, scientific thinking is not about absolute certainty. The conclusions drawn from scientific research are probabilistic- generalizations that are correct most of the time, but not every time. People often weight anecdotal evidence more heavily than probabilistic information. This is an error in thinking, leads to bad decisions, and often, irrational thinking. It is important to accept statistical predictions aren't perfect. These predictions are based on samples (groups, categories intending to represent populations) and will be correct more often than not.


To learn more about statistical thinking refer to - In Evidence We Trust   

Sunday, September 15, 2019

The Illusive Scientific Method


Science writers, educators and sometimes researchers have a tendency to refer to the scientific method. This method is the systematic process, methodology, used in science to acquire knowledge and test claims. One huge problem with the referral to a singular methodology, it doesn't exist. Science makes use of multiple methods, models and inferential strategies. Why so much discussion about and illusory method? Maybe, it demonstrates and appeal to tradition, an appeal to popularity, a misunderstanding of the sometimes complex nature of science and its relevant components and interacting factors, or mentioning a singular scientific method is easier than the alternatives (demonstrating cognitive miserliness- which is normal for most).


Science is broad; it consists of many components and sub-components. Discussions regarding science are sometimes short-circuited by discussing a single component. These types of discussions oversimplify the wide range of science, its development, and implications. A full appreciation of science requires much more than a focus on a singular element. Skepticism is an element of a scientific attitude and is important, but a skeptical attitude alone—without other cognitive skills and knowledge—doesn’t make one a scientific thinker. Science is all about skepticism, so say the popularizers of science. Skepticism is important, but without the knowledge and appropriate skills, this characteristic will not make one a scientific thinker. Science is hard. In the words of Albert Einstein, “Things should be made as simple as possible, but not any simpler.”  Science: The Vast Enterprise


Scientific literacy, in the context of this article and my research, is synonymous with general scientific knowledge. Scientific literacy in this form involves remembering scientific facts, theories, principles, and so on—products of scientific inquiry. This form of literacy is sometimes referred to as a form of derived scientific literacy. Scientific literacy is important, however other science related concepts are just as important. Scientific cognition is not the same as scientific literacy.

Scientific cognition (thinking) involves complex cognitive mechanisms. Scientific cognition involves much more than general scientific knowledge, procedural skills to conduct research, attaching “science says” to your statements, a science degree, perpetuating views of popular science figures, identifying yourself as evidence based, asking for evidence, being skeptical, etc. Scientific thinking involves thinking that can also be used out of the lab. At the very least scientific cognition involves philosophy of science, scientific methodology, quantitative reasoning, probabilistic reasoning, and logic (deductive and inductive). To learn more about scientific cognition and scientific literacy refer to- ScientificCognition and Scientific Literacy at Kentucky Academy of Science 

A better understanding and promotion of science is needed. Science does not involve a singular method. Rethinking science education involves at the most basic level acknowledging the broad nature of science, promoting the message that science is hard but learnable, identifying and accepting the limitations of science, and understanding science is about more than just retrieving scientific facts from memory. Science education is valuable for everyone. Recommendations for future science education: Rethinking Science Education   

Precision in language is essential for science communication. The misuse of the term "scientific method" is one of many examples of unclear language. In my book- In Evidence We Trust- I point to examples, often seen in textbooks and used by instructors, to the lack of precision in scientific language.  That is a problem.  Clear communication requires precise, consistent use of terminology.  Students often accuse me of being a semantic stickler.  I accept the label.  When talking and writing about science, practice precision and consistency.