"Making Statistics Make Sense with Web Applets"
with Jonathan Lee, Western University
Hosted by: Leigh Slauson, Capital University
Interactive web applets are often advocated as an engaging way to enhance students' comprehension and sense of mastery of the material. We have designed activities for use in an introductory statistics course on measures of centre and spread as well as confidence intervals and p-values based on the use of applets available online. These short applet activities have been set up in a learning management system and they would be appropriate for any type of course ranging from traditional classroom-based to hybrid to fully online. We also conducted an experiment to investigate the applet activities' effects on student attitudes and comprehension. In this presentation we will share some of the applet activities and highlight some preliminary results of our experiment.
Having trouble viewing? Try: Download MPEG-4 Podcast (.mp4)
Comments
Michael Granaas:
Your sense of competence data is consistent with the Dunn-Kruger effect. An effect where unskilled folks over estimate their competence on a subject. Initial over estimation of skill seems likely to account for your observed decrease in sense of competence.
I am delighted to see people collect actual data--tech solutions have taken on a magical aura and are assumed to be beneficial unless proven otherwise. Data like these challenge that aura.
I did not notice if you had any sort of ability pre-test. That seems a likely covariate to reduce error variance.
Jonathan Lee:
Thanks for the comment Michael, it's very insightful and helpful. Regarding your pretest question, we did not have have an explicit pretest aside from the survey but the test performances may be able to be viewed as a pretest for the final or those students that accessed the web applets they were not exposed do right before the final when all applets were made available.
Janet Sanders:
Interesting study. Would you share the applets that you used?
Bethany White:
Thanks for your great question, Janet.
We selected the following applets to use - we didn't develop new ones, just chose some that suited our learning objectives from those available online:
Centre & Spread - web.uvic.ca/~esfchan/stats/Descriptive%20Statistics%20Applet/descriptive.html
Confidence Intervals - www.math.usu.edu/~schneit/CTIS/CI/CI.html
p-values - www.amstat.org/publications/jse/v16n3/PValueApplet.html
The activities geared around use of the applets would play out as follows:
At the beginning of their lab sessions students would be shown a brief screencast introducing the applet and demonstrating how to vary things. Then they would each complete a short assessment (about 8 questions) through their learning management system (WebCT). The first couple of questions would be completed before students played with the applet. Then they were guided to vary certain things in the applets to complete subsequent questions. The activity wrapped up by asking them an open-ended conceptual question about the concept being explored.
Once we have all the data and have completed the analysis we'll write up a paper on the results and provide specific details on each of the applet activities (the questions and the screencasts). At that time we will also make these available online.
Thanks again for your question,
Bethany