|
|
| by
StefanWaner and Steven R. Costenoble
|
| |
|
 Ratings
|
|
|
| Reviewed: |
Sep 08, 2010 by Statistics Editorial Board |
| Overview: |
The material covers confidence intervals for large and small sample sizes using both the z and t distributions. |
| Learning Goals: |
The learning goals of this site are to:
*understand the benefit of sampling distributions
*calculate confidence intervals for the population men in several different situations
*accurately use the information attained to make inferences on a population using confidence intervals. |
| Target Student Population: |
Introductory-level students. |
| Prerequisite Knowledge or Skills: |
Basic understanding of the Normal distribution, standard deviation, and reading both Z and t tables would be helpful. Perhaps an introduction to confidence intervals in class ahead of time, too. |
| Type of Material: |
Tutorial |
| Recommended Uses: |
Out-of-class additional practice for individuals. |
| Technical Requirements: |
Browser—worked well in Firefox. A calculator is needed for the confidence interval calculations. |
|
|
|
| Strengths: |
It was made into a dialogue which answers many questions students have. Also, step by step instructions allowed the reader to easily follow along. This is a friendly and easy-to-follow explanation of important confidence interval concepts. The examples that allow students to try problems and check their work are very good. |
| Concerns: |
The graphics on the mathematical symbols are not always clearly readable. It may help to explain some of the graphics in a bit more detail, namely the first set of charts. There were many misspellings and grammatical errors throughout the pages as well as the answer keys. In some instances, although just typographical errors, the information was not correct: for example, the sample mean will not be within 2 standard deviations of n, but rather the population mean. Can you do a 1-sided hypothesis test with a CI? Our book says you can only use it for 2-sided hypothesis tests. (End of the tutorial.) |
|
|
Potential Effectiveness as a Teaching Tool |
Rating:     |
|
| Strengths: |
There are many examples presented and step by step instructions for how to complete the problems were laid out which allowed for good practice. This module covers many types of confidence interval situations and gives examples of each. |
| Concerns: |
I would like to see a chart summarizing the different procedures for different setups. Large sample, normally distributed, pop. st. dev. known vs. small sample, normally distributed, pop. st. dev. unknown, for example.
Some graphics to go along with the calculations of 20 confidence intervals would have been helpful. Discuss how the width of the confidence interval doesn’t change, but the center does. Perhaps show similar graphs for other confidence levels to show how the width does change.
I think the students might confuse sample size with number of samples, so it would be good to encourage them to change the number of samples and explain the difference in these concepts a little more.
Encourage them to press the sample button multiple times so that they can see the results aren’t always 18/20 “yes”—confidence refers to a method that gets it right 95% of the time in the long run.
The idea of needing a confidence interval was addressed, but it was taught in a way that students were simply plugging numbers into a formula rather than having them discover a confidence interval on their own. Perhaps a bit more reliance on their knowledge on the Normal distribution would allow more critical thinking on the student’s part. |
|
|
Ease of Use for Both Students and Faculty |
| Rating: |
|
| Strengths: |
There were options to both check and see answers for all questions the students were asked to participate in. There were several colors used so as to be more visually appealing. |
| Concerns: |
Perhaps using a non-serif font would aid in readability. Although some graphics did help in understanding, I think a variety of pictures would help: for example, instead of giving a list of 20 different confidence intervals, so actually show the confidence intervals in relation to a normal distribution may help. In the student calculations, perhaps a “hint” button could remind students what the letters mean or that alpha should be listed as a decimal instead of a percent. Example 3, there is a typo in the word “adjustments” in the 1st line. |
|
|
