The two worksheets enable instructors to demonstrate how changes in the magnitude of the treatment effects and of the standard deviation of the error term will impact significance in a One-Way ANOVA model. The user specifies three input values that influence the simulation of random observations. ANOVA calculations are provided for the student, leaving the focus on the interpretation of the results. The mirror site (found at http://misnt.indstate.edu/cmclaren/ANOVA_Note.doc) contains an article that can serve as a teaching note to accompany the worksheets.
Content Quality Concerns:
This Excel simulation compares the means of three groups but no context is given for this comparison. It might make more sense to students to put this into a context to begin with. Related to this, there is no discussion in the tool or within the teaching notes about issues related to effect size or practical importance. The graphs are a nice addition to the tool, but I think it would be better to include these near the top of the tool, perhaps even above the ANOVA table, so students can more easily see what happens to the groups as they make changes to the treatment effects and error. When the tool first opened on my computer, I could not see the graphs right away and had to scroll over to them. At first, my concern was how the actual data were generated using only N(0,1) error terms even when the user changes the standard error. By clicking on the cells, I saw the equations and understand that the calculations are correct for generating the three samples.
Content Quality Strengths:
The strength of this activity is that it allows students to observe the effects of changing the parameters on not only the estimates of the parameters, but on the test statistic and p-value. One can see that with a lot of variation, small treatment effects often are not detected. This tool encourages students to make and test conjectures about treatment effects and error and how these affect the F-statistic within the context of ANOVA. The tool includes not only a data set and an ANOVA table but graphical displays of different groups. The tool uses accepted notation and includes notes for the user.
Ease of Use Concerns:
Although there are some notes and comments included in the spreadsheet, the notes do not seem to include very clear directions for how students might use the tool. It seems like it would be hard for students to use this without some prior instruction or direction from instructors (and students who may not be familiar with Excel may not realize the need to drag the mouse over the red tabs to see the comments). The graphs included off to the right might work better near the top of the tool, and adding more color to these graphs might make the entire tool more visually appealing.
Ease of Use Strengths:
Any student who has access to Excel can use this, and students only need to type in three values in order to perform a simulation. Students can readily see the results of their manipulations on the screen. The screen is not too cluttered with information and all tables are well labeled.
Potential Effectiveness Concerns:
This might work well as a demonstration in class, or as a guided classroom activity (in a computer lab), but it's not clear how this might work outside of class (i.e., would students need very structured questions/instructions to be able to work on this at home?), or what kinds of assignments could be written for use with this tool.
Potential Effectiveness Strengths:
The tool allows students to experiment and to explore, and it allows them to see how certain factors affect the magnitude of the F-statistic. Students can make conjectures and immediately test them and get instant feedback, and this tool has the potential to lead to very rich discussion on the mechanics of ANOVA. Rather than just seeing an ANOVA table as a set of numbers, students will hopefully realize what these numbers refer to and how variability within and between groups affects these numbers. The teaching notes are a nice addition to the Excel simulation, and I particularly like the inclusion of teaching tips and classroom discussion points.
Potential Effectiveness Rating:
Source Code Available:
Source Code Available
Intended User Role:
Free for Nonprofits