Confidence Intervals

Overview

One of the most important questions for an instructor, particularly one teaching a course dealing with quantitative methods, is whether or not students actually understand the material. It is one thing to memorize and perform a series of calculations, and another thing to know what these calculations mean in a real world sense. In previous biostatistics courses, students have particularly shown trouble in understanding the concept of confidence intervals and their applications.  Even if they’re able to perform the arithmetic necessary to create one, many do not understand a confidence interval’s significance.  It is this deficiency that inspired the creation of this module, which is aimed at helping students understand what a confidence interval is in both mathematical and pragmatic terms.

The design of this module takes a strong multimedia approach, using an animated lecture, interactive tools, and video problem solving to give students another means of visualizing the course material.  Not only are students given examples on how to solve problems involving confidence intervals, they are given additional tools to understand how to define such problems as well.

The animated lecture was created using Flash, and uses a very clear and simple design to convey the key definitions and illustrative examples.

VIEW FULL SCREEN

The simulation gives students an opportunity to see an illustration of a confidence interval, helping make a relatively abstract concept more concrete:

VIEW FULL SCREEN

Three videos provide worked examples, known to be a highly effective tool for helping students learn problem-solving skills. One example is shown here:

Finally, the module includes an assessment tool for students to evaluate their own understanding in a low-stakes environment:

VIEW FULL SCREEN

The reason for taking a multimedia approach is the success it has met with in earlier implementations. There is a wealth of contemporary literature by educational researchers that supports the hypothesis that multimedia-based education helps improve learning. Furthermore, based on my own learning experiences, and the consideration that different individuals learn in different ways, I thought this could be a viable solution to the issue of student miscomprehension.  Materials were designed to be short, less than 10 minutes in length, in order to maintain student attention, and to allow intake of the material in manageable sections.  Questions were designed to probe for common misconceptions on the subject of confidence intervals, and to help students identify and divorce themselves of them.

Assessment

To assess whether or not this module completes its aims, student performance was compared to previous semesters that did not feature this module.  In particular, an ABET test question that was used both during the module semester and in previous semesters was used for comparisons. Interestingly, students in the two semesters did not show a consistent difference in their ability to “plug and chug.” Students in the semester that implemented the module, however, showed strongly improved understanding of how to interpret confidence intervals—that is, showed improved comprehension of the targeted concepts.

In addition, students generally found the module helpful. The students were asked about the value of each of the elements of the module for helping them understand confidence intervals:

Question 1: overall assessment of online activity
Question 2: animated lecture
Question 3: confidence interval simulator
Question 4: solved example problems

More than 70% of the students found the module to be “a good amount” or “a lot” helpful to their understanding, and at least 50% of the students identified each of the elements of the module as falling into these categories:

In addition, we asked whether the online practice assessment was helped students identify their misconceptions about confidence intervals, and >85% of them indicated that it did (Question 5).

View the audio/PowerPoint presentation by Zane Ricks about his BOLD project.