MERLOT - Multimedia Education Resource for Learning and Online Teaching
Home Communities Learning Materials Member Directory My Profile About Us

Peer Review

Become a Member | Log In

VassarStats: Website for Statistical Computation

by Richard Lowry
 

Ratings

Overall Rating:

4.5 stars
Content Quality: 4.75 stars
Effectiveness: 4.25 stars
Ease of Use: 4.75 stars
Reviewed: Jan 30, 2011 by Statistics Editorial Board
Overview: This is a powerful tool for students and teachers to have available to use, both in classes and for problems. This material provides interactive calculators for various statistical techniques, tests, and applications.
Learning Goals: This is a toolkit to calculate many different statistical results for inferential statistics, tests, etc. As such, it is a tool for learning, not a lesson with a goal. The purpose of this site is to provide opportunities to compute statistics on user inputted data.
Target Student Population: Students in elementary statistics courses who are learning inferential statistics.
Prerequisite Knowledge or Skills: Minimal prerequisite knowledge is needed, as concepts are explained in great detail, but would be most helpful to use concurrently with a statistics course or statistical background.
Type of Material: This material is a huge collection of a vast array of statistical concepts, calculators and overview.
Recommended Uses: This site would be very helpful for in class data collection, lecture aids, individual learning and homework use.
Technical Requirements: This material requires an internet browser.

Evaluation and Observation

Content Quality

Rating: 4.75 stars
Strengths: The material is presented clearly, the values I put in were calculated correctly, and the results were clearly labeled. Once one knows the format, it was easy to see the values, and the labels told what the value was. Several different results were given, for the different types of test intervals, so it would be very useful. The site looks professional, with minimal typographical errors, has a consistent theme, and link/pictures/help are all clearly labeled.
Concerns: Only a small weakness: All the information was not on the first screen without scrolling, so I was temporarily confused about how to get the results out. Just a little sentence saying, scroll down for results, or something similar, would be helpful to the first time user. After that, it was great! It might be helpful if the area shaded into the curve was available in picture form for the z to P calculator. And it is unclear if “one-tailed for –z” is meaning that the alternative hypothesis is “<” and “one-tailed for +z” is meaning that the alternative hypothesis is “>”. In which case the probabilities should add to 1, instead of being the same. If this is not the case, I think it should be specified. In the “Convert to standard scores” tab, procedure steps, the “When all” bullet point is not in the same format as the others, and the word “all” is repeated unnecessarily. On the simple graph maker, 5 columns are allowed, but only 4 colors. It might be helpful to add another color. In the Distributions tab, there seems to be an extra bracket after “Frequency Data”. For Logistic Regression, there are 3 input boxes at the bottom of the page that seem to be out of place. For t-Test for independent and correlated samples, the results are the same no matter which selection (independent or correlated) is chosen. Output for .95 and .99 Confidence intervals for the estimated mean of a population is a text output in a new window rather than specified outputs in the same window which has been the norm for most of the site. Two-Factor ANOVA with Repeated Measure on Both Factors, the Procedure, does not have bold bulleted headlines like other Procedure steps. For ANCOVA, also the procedure steps are not as clearly labeled as in previous pages.

Potential Effectiveness as a Teaching Tool

Rating: 4.25 stars
Strengths: This is a tool, not a set of learning modules, so this is not evident here as much. I did find the presentation of the different test conditions at once to be helpful to students. The material develops and explains in great detail a multitude of statistical concepts and allows users to input data freely and easily. The toolkit could be easily implemented into a course, and since it presents the results after the conditions are entered, it could be easily used by students. Using this tool, the students won't get lost in the process of looking up values in the table, etc., which can happen easily otherwise.
Concerns: It is debatable whether opportunities for student learning are lost when calculations are simply given to students according to the user input; however, students do need to know how to use such tools effectively to achieve the answers they are looking for. It may be helpful to be more explicit as to what computations are used to derive the output given in many of the situations. Formulas may not necessarily be used by the students, but it may be helpful to see what formulas are used.

Ease of Use for Both Students and Faculty

Rating: 4.75 stars
Strengths: It is easy to use, and after I found the results, was great! It's not too cluttered, but the figures one would use are readily available when clicking, so they don't clutter up the screen, but are accessible. The site is very visually appealing, with large consistency and clarity throughout. Instructions are given to make it clear what user input is needed and what output will be given. Help menus are available for further explanations of concepts and computations behind the scenes.
Concerns: It is not always evident what areas the user should not input data. There are some pages that have yellow output boxes indicating the computer will use those spaces, whereas other pages, the output boxes are white and show no indication that the user should not place data in such boxes. Some boxes also have dashed lines rather than a yellow highlight, so slightly more consistency might be beneficial. Also, again, the results need to be scrolled down to. A simple indication that this was needed would help the first-time user.
--%>