This applet simulates the creation of a confidence interval for a given sample size and confidence level. Users can select from a normal, skewed, skewed binomial or custom created population. Sample size can be selected to be between 6 and 40. Confidence levels can be selected to be 90%, 95% or 99%. Additionally, the calculation for the confidence interval can also be displayed.
The student can visualize how a confidence interval either captures the true population mean or does not and how the performance of the interval depends on the type of population, sample size, significance level and variance.
Target Student Population:
Introductory level statistics students.
Prerequisite Knowledge or Skills:
Students should understand descriptive statistics and sampling distributions before using this applet.
Students should understand what a population is (and types of distributions such as binomial and normal), students should also understand what a confidence interval for the mean is in general and the concepts of sample size and confidence level.
Type of Material:
Recommended use is as an in-class activity when introducing confidence intervals for the mean so that students can determine how changing different aspects of the margin of error affects the confidence interval.
It can also be used as an activity to show what is meant by the confidence level since students can clearly see how many confidence intervals “miss” the true value of the parameter.
There is substantial guidance and examples provided for using the learning object, thus it is appropriate as a out-of-class learning tool as well.
Java required. Works fine with Chrome, Firefox, or IE.
Evaluation and Observation
This applet is a great in-class or out-of-class tool for teaching a correct conceptual understanding of frequentist confidence intervals for a single population mean. The applet provides limitless variations on the population so the instructor or student can suit it exactly to a particular setting. It is easy to use and the graphics are fantastic. It really shows exactly what is going on with the confidence interval when certain aspects are changed such as confidence level and sample size. It also nicely displays how differing samples produce different confidence intervals.
The only drawback is that the number of samples from the population max out at 20. You can of course run this several times, but it would be great to have an option where you can sample the data 100, 1000, etc… times and see the relative frequency of catching the mean.
Potential Effectiveness as a Teaching Tool
This is a great tool for teaching confidence intervals. It’s easy to use and doesn’t require any advanced knowledge. Simulations can sometimes be difficult to use because they require some computing skills, but this is easy enough to use for anyone with or without computing skills. Ease of use is definitely one of the most outstanding features on this applet. Also, the range of options are outstanding giving this applet the potential to be a very effective learning tool for students.
Excellent and useful tool. No big concerns!
Ease of Use for Both Students and Faculty
This is a very usable applet. It is fairly intuitive to run. The instruction page is very detailed.
The only big concern is that the sample button must be selected for each individual sample of the data. Depending on how one plans to use this applet, it could become time consuming to click the sample button many times. Also, the instruction page seemed a little too detailed and students may feel overwhelmed if asked to read it!
Other Issues and Comments:
The only other comment that I’d make is to include some higher sample sizes such as 50, 75, 100. This may not be feasible in that it’d make the applet run too slow, but it would be nice to have a few higher ones to work with.