In an early post, I stated that I “teach naked.” (follow this link to the post: https://statisticalsage.wordpress.com/2010/08/25/teaching-naked/) That means I do not use PowerPoint slides when teaching statistics. Yes, I made a set of PowerPoint slides, as the editor of the stats book made it clear as a requirement for publishing our text http://www.pearsonhighered.com/educator/product/PowerPoint-Presentation-Download-only-for-Statistical-Concepts-for-the-Behavioral-Sciences-4E/9780205626281.page, but I don’t use them. Several of the sages and I concur that teaching statistics is not benefitted from the use of PowerPoint slides, at least not when we are teaching it. Though there are more ways to integrate technology into the teaching of statistics. Earlier this week, we were visited by Math Professor Livie Carducci who provided us with some specific information on how to implement visual demonstrations through technology to aid in conceptual understanding https://statisticalsage.wordpress.com/2011/05/24/using-mathematica-deomnstrations-to-visualize-statistical-concepts/. Of course, many of us have to teach students about how to use statistical tools for calculating large data sets.
Yet others (e.g., Shaltayev, Hodges, & Hasbrouck, 2010) feel that technology should take center stage in a class like teaching applied statistics, and that having students use paper and pencil to learn about the concepts of statistics is not beneficial. Before I continue, I have to say, I agree that calculations are not worth students time if students are being asked to use computational formulas, as the act of calculating them will not help them to see what the statistic is doing. Though, I do feel there is a benefit in having students calculate statistics by hand if they are using definitional formulas for calculating small data sets.
Not all calculations will be as valuable to students. Having the data sets out of context of answering a question certainly doesn’t aid to students conceptual understanding, as well as keeping statistics in concrete context. Specifically, we should start with a question and end with an answer through the application of statistics to assure maximum understanding. This is optimal for having students truly comprehend what statistics can tell us, and what they can’t. Having these questions as “real” and relevant as possible is most helpful (here is a pitch for service learning projects as a means of providing real context for students).
Yes, I still lean toward being less plugged in when teaching, but as I stated in prior blogs, there is more than one way to teach, and being aware of differing methods is one way for us to improve our teaching. Shaltayev and other contend that by having students focus upon understanding the question being answers, selecting the right statistic, and interpreting the results instead of working on calculations, students will better understand statistics. I certainly would agree with this statement, where I differ with Shaltayev and others is I do not see hand calculations and conceptual understanding as mutually exclusive.
So … start with a question. Select the study and statistics, “gather” a small data set, and hand calculate it using the definitional formula, then make a decision and interpret the results. The entire context is what is critical, and calculating the definitional formula is part of that larger context.
By using the definitional formula students can see what the number is telling them, as they can see the standard error in the t-test. They can see the effect going into the numerator of the F-test, but that error is mixed in with the effect. By hand calculating a correlation coefficient, students can see how when two variable covary how that will impact the numerator. Though some well placed questions (Socratic Method) coupled with constraining the situation, having students master component pieces before integrating it all together, will truly take students to a deeper understanding of statistics.
and the accompanying video http://chronicle.com/article/Video-A-Professors-Plea-/48404/
Shaltayev, Dmitriy S., Hodges, Harland, & Hasbrouck, Robert B.(2010). VISA: Reducing Technological Impact on Student Learning in an Introductory Statistics Course. Technology Innovations in Statistics Education, 4(1). Retrieved from: http://escholarship.org/uc/item/1gh2x5v5