## Using StatCrunch For A Final Exam

*May 20, 2010 at 5:02 pm* *
1 comment *

This semester I am giving my Intro Stats final exam in a computer lab. The exam is completely inferential covering confidence intervals and 10 hypothesis tests – 1 mean z, 1 mean t, 1 proportion, paired difference, 2 mean, 2 proportion, 2 variance, goodness of fit, chi-square independence, and ANOVA.

My students will be using StatCrunch to generate confidence intervals and p-values. On the last day of class I allowed my students to create a “StatCrunch cheat sheet”. They were able to write down important steps for finding the appropriate test in the StatCrunch menus, as well as important facts to remember for particular tests such as remembering to uncheck the Pool Variances option on a two mean test. I will give them these cheat sheets when they arrive to take their final.

### What I Am Testing

I’m really happy with my approach with this exam, and this course in general. Here are the skills that are being tested.

- Upon reading a problem, a student must be able to determine which statistical test is appropriate.

This skill is very valuable in their statistical future. Whether it is in a future course in their major or out in the workforce, the first step in solving a problem with statistical analysis is to determine which test is appropriate. - A student must be able to extract all important information and enter that information into our statistical analysis package – StatCrunch.
- A student must be able to interpret the results generated through technology.

Again, this is a very important skill. Students will often find themselves in a position where data will be presented to them and they will have to interpret the results. In my experience, students will find that they will be presented others’ results more often than they will have to perform their own analysis. They must be able to read, understand, and interpret the results of others. - A student must be able to correctly interpret a confidence interval or the results of a hypothesis test.

I ask questions that ensure that my students completely*understand*what they are doing rather than just mindlessly mimicking steps. It is here that I am able to focus on conceptual understanding.

### Perceived Benefits

My students have not devoted their mental energy to figuring out how to evaluate a particular formula. StatCrunch has allowed them to spend more of their energy on truly understanding the material in the course. The course is now taught in a way that has real world value.

The switch was not easy for me. Sometimes I would wonder whether this approach was valid because my students were not doing many calculations by hand. Confession – I used to have my students do ANOVA calculations by hand. This semester I showed them all of the formulas, and they were unanimous in their support of our new approach.

The final exam is next Tuesday and Wednesday. I’ll be back to share the results, as well as the results of a student satisfaction survey I’ll be giving.

### Your Input Is Requested

Do you use StatCrunch, or some similar technological tool, in your statistics courses? Would you like to share your approach, and particularly your exam strategies, with other instructors? Do you have any questions on StatCrunch? If so, please leave a comment or reach me through the contact page at my web site – georgewoodbury.com.

-George

*I am a math instructor at College of the Sequoias in Visalia, CA. One of my Thursday topics is Technology in Education, including StatCrunch. If there’s a particular topic you’d like me to address, or if you have a question or a comment, please let me know. You can reach me through the contact page on my website – http://georgewoodbury.com.*

Entry filed under: General Teaching, StatCrunch, statistics. Tags: announcement manager, anova, cheat sheet, chi square, college, confidence interval, education, F test, final exam, formula, george woodbury, goodness of fit, hypothesis test, independence, intro stats, Math, math study skills, mean, no formula approach, p value, p value approach, Pearson Education, proportion, StatCrunch, statistics, stats, t test, teaching, technology, variance, woodbury, z test.

1.Sheila Pisa | May 21, 2010 at 7:15 amHi George,

A colleague forwarded me your blog post about StatCrunch because I am using it this semester. I am also giving an all StatCrunch test on hypothesis testing, correlation, and regression – but it will be on June 2.

For the final in this class (June 9), I have made five groups and they each decided on at least one research question. They then designed a survey. I reviewed the surveys with them this week and we discussed data collection methods and how they would determine if their sample was large enough. They will be administering their surveys for the next week. Once they’ve collected and analyzed their data, they will do a half-hour presentation for the class. Their presentation must include descriptive as well as inferential material and they must use visuals such as charts, handouts or powerpoints. Each member of the team has to write about how they divided the work and to grade each other member’s effort relative to their own. The students observing the presentation will also be allowed to submit a grade for the presentation.

I have never done this with a class before, so if you have any experience or could provide any advice, I’d appreciate it. I am also working with other faculty who teach stat to coordinate our stat classes so that we are doing similar projects. Everyone is interested so I am looking forward to seeing your test results.