What happens in a statPREP workshop?
statPREP workshops have three main goals:
- Provide participants with the tools and techniques needed to work with data and to bring data-centric teaching into the classroom.
- Demonstrate how data-centric approaches can be used to teach statistics in a better way.
- Establish contacts among the participants and with the workshop hub leader to provide ongoing support as participants merge statPREP resources into their teaching.
Tools and techniques
Working with data involves both computational resources and conceptual approaches to data. The first half of each workshop will provide instructors with both of these.
We use a computational platform that is free and open-source, and we provide ongoing access to you (and your students, if you wish) in a way that requires no software installation on your or your students’ computers. The platform integrates narrative, calculation, display, and assessment in interactive tutorial format. During the workshop, participants work with the statPREP workshop leaders using tutorials designed to introduce techniques for working with data. (Similar tutorials can be used as part of classroom demonstrations and, if you like, as activities/homework for your students.)
Using the workshop tutorials, participants will acquire a compact but complete collection of computer commands for data analysis, visualization, statistical calculations, and simulations. These commands draw on extensive experience in teaching statistics with software as well as software interfaces for data science. The commands are designed to be easy to learn and easy to use. There are two simple constructs at the heart of the approach:
- “Tidy” data tables for storing data and for communication between successive steps in the process of data preparation, analysis and display.
- A modeling notation for expressing the relationships that are to be examined in data and for constructing rich displays of data.
Using these two constructs systematically allows a large variety of computational tasks to be performed using simple, standard command templates. Under resources there is a link to tutorials that are presented at the workshop, but you can also preview the material by clicking on this link for a brief introduction.
For those interested in the technical details of the statPREP computational platform, here some further information. We use the R language together with a computational ecosystem largely developed by RStudio. The interactive tutorial documents are written in RMarkdown, and can be displayed and worked with in an ordinary web browser. Commands for statistical calculations and graphics are based on ggplot, dplyr, and mosaic. Ggplot is one of the most widely used graphics systems in professional data science. Dplyr is a similarly widely used “data wrangling” system, often used by professional data scientists. Mosaic has a track record of several years in simplifying statistical computing for use in the classroom. StatPREP has integrated ggplot and dplyr with mosaic, providing a remarkably easy-to-learn syntax. A major advantage of the tutorial format is that commands can be presented with any level of scaffolding that’s desired, from push-the-button to fill-in-the-blanks to write-your-own.
StatPREP is also providing access to the tutorials using a web cloud service. But all of the software can be easily and freely installed and run on an ordinary laptop computer. This will be particularly useful for those instructors who choose to modify materials provided by statPREP or to write their own lecture notes and student activities.
As described above, the workshops start with tutorials that introduce computational, graphical, and statistical techniques to instructors.
The second half of each workshop is about using those techniques to enhance statistical teaching and learning. To be sure, the software is perfectly capable of the calculations encountered in statistics courses, so instructors could use the platform for providing student access to statistical software. But the statPREP lessons will emphasize how working with data can be used to improve teaching and enhance student understanding of statistical concepts through compelling demonstrations and activities.
In the workshops, we’ll start with course lessons written by statPREP leaders. We will then go into more detail of the 6 major topics of a statistics course. These topics are Descriptive Statistics / EDA (exploratory data analysis), Normal Curve, Hypothesis Testing, Confidence Intervals, Regression, and Sampling Variation.
StatPREP is committed to continuing to support and work with instructors after the workshops. We will help you customize lessons and demonstrations to suit your course and to use the software to express your own ideas. As part of the follow-up, regional hub leaders have been appointed to provide a face-to-face point of contact, and to support contacts between the participants.
Over the five-year course of the project, several hundred instructors will receive statPREP training. StatPREP national leaders and hub leaders will work with that community to develop and refine lessons and to support instructors in supplementing additional parts of their courses with data-centric teaching.