Abstract:

Data is literally increasing at an exponential rate, but will our students be able to use it? This question is vital to business and industry, and its answer depends on our ability to produce a large, highly skilled workforce. Despite the growth of academic data science programs, the supply is not meeting the demand. On top of this is the need to produce a data-literate population. As the Covid pandemic taught us, even a significant portion of a population that is not data literate can significantly impact a country’s economy, social cohesion, and trust in its institutions.

These two parallel issues of training a workforce of skilled data scientists and a data literate population have a common denominator: inference. Inference is a foundational part of statistics; it is the first guideline of the Guidelines for Assessment and Instruction in Statistics (GAISE) report, “to teach statistics as an investigative process of problem-solving and decision-making.” However, how well are we teaching the process of inference? Our study indicates that students learn the steps but not the logic of the inferential process. Our study supports the results from other larger-scale studies that show that even with new innovative methods for teaching inference, such as the Statistical Based Inference (SBI) method, gains in student learning are either mixed or modest.

These findings led us to a second question: how can we improve how we teach the logic of inference? We propose a new approach called “Question, Explain, Do” (QED). Unlike other methods (such as “Tell, Show, Do” by DeVeaux and Velleman or the “practicing connections framework”), QED is based on George Cobb’s recommendation to “put the logic of inference at the center of our curriculum.” It is an organizational principle that centers the entire curriculum about the logic of inference by redefining statistical reasoning, thinking, and literacy. We explain the principle underlying QED and sketch how QED can be used to help students see “The Big Picture of Statistics.”

Keywords:

Statistics education, inference, teaching.