"An overview scheme for introducing statistical inference by labeling inference actions"
Frank J. Matejcik, South Dakota School of Mines & Oglala Lakota College
Abstract
What is the best way to introduce inference? This issue was actively debated in early January 2013 in a series of email exchanges between members of the ASA Section on Statistical Education. In response to this discussion William Notz of Ohio State University lead a Roundtable on this topic at the Joint Statistical Meetings in 2013. The focus of these discussions was should confidence intervals or hypothesis tests be presented first, and whether proportions or means should be considered first. Some students, particularly engineering students, first see inference in control charts, therefore this form of introduction should be considered, also. Mentioning control charts at an introduction to inference in Statistics helps students that will see control charts in engineering and business classes. Inference topics can be seen as disjoint detailed procedures, which are difficult for global learners. The structured overview presented here should be helpful for global learners. This overview of statistical inference discusses four uses which are to adjust (control charts), to measure (confidence intervals), to excite (the p-value method of hypothesis tests), and to decide (the critical region method of hypothesis tests). The acronym AMED is used. Preferences for the forms of inferences are mapped to Aristotle's four temperaments. This mapping suggests the AMED presentation of inference may motivate many types of students. The name "Amed" means "greatly praised", which might be a good way to think of inference. While the AMED presentation may seem that it would take considerable class time or space in a text, this video presentation of AMED vimeo.com/86650563 takes just under eight minutes. And, the video has some playful additions, too.
Materials
- Download slides (PowerPoint)
Recording
(Tip: click the fullscreen control)
Having trouble viewing? Try: Download (.mp4)
(Tip: right-click and choose "Save As...")
Comments
Nicholas Horton:
The paper by Wild et al (new.censusatschool.org.nz/resource/building-inferential-reasoning-in-statistics) is another useful resource. We need to better understand how to introduce inference to our students, and this poster raises some helpful suggestions. @askdrstats