Using Group Projects in a Business Statistics Course

Using Group Projects in a Business Statistics Course: Experiences from the Past and Possibilities for a Future Course Joan Donohue University of South Carolina 1

Topics Currently Covered in our Introductory Business Statistics Course 1. 2. 3. 4. 5. 6. 7. 8. Descriptive Statistics Probability Discrete Probability Distributions Continuous Probability Distributions Sampling Distributions Confidence Intervals for Means and Proportions One-sample Hypothesis Tests for Means and Proportions Two-sample Hypothesis Tests for Means and Variances (a group project is done at the end of the semester) Starting Fall 2017, the above course will be taught by the Statistics Department (except the project in #8). A new required Business Statistics course will be taught in the Business School, covering more advanced topics and including ~3 projects. 2

Potential Topics for the New Required Business Statistics Course 1. One and Two Sample Hypothesis Tests - brief review 1 st group project – likely the same one used in old course 2. ANOVA – one and two factor 3. Simple Linear Regression 4. Multiple Linear Regression - 2 nd project 5. Time Series and Forecasting 6. Chi-Square Tests 7. Non-parametric Tests 8. Quality Control - 3 rd project topic? Questions: Use 2 -person groups for all projects? Let students work in the same group for all 3 projects? Topic for 3 rd project? Should we require students to purchase Mega-Stat or other statistical software instead of using only Excel? 3

Excel Data Analysis Tools Anova: Single Factor Anova: Two-Factor with Replication Anova: Two-Factor Without Replication Correlation Covariance Descriptive Statistics Exponential Smoothing F-test Two-Sample for Variances - Project 1 Fourier Analysis Histogram Moving Average Random Number Generation Rank and Percentile Regression - Project 2 Sampling t-test: Paired Two Sample For Means t-test: Two-Sample Assuming Equal Variances - Project 1 t-test: Two-Sample Assuming Unequal Variances - Project 1 z-test: Two Sample For Means Chi-square, Non-parametric tests, and Quality Control are not included. For Time Series, Exponential Smoothing and Moving Averages could be used. Mega-Stat and other software packages cover all of these topics. 4 4

Current Project Involving Twosample Hypothesis Tests Students work in groups of 2 and can chose their partner. Otherwise I assign them a partner (usually someone they sit nearby to). Typically the partner situation works out very well – almost no complaints! 5% Proposal Develop a hypothesis that compares two groups of USC students in terms of some characteristic that can be addressed with a 4 -item Likert-scaled survey. 20% In-class Presentation Give a 5 -6 minute Power. Point presentation in class. 75% Written Report Write a paper that explains all aspects of the study (appendices include your survey and Excel worksheets). 5 5

Examples of Hypotheses Submitted by Students who are the first-born child in their family are more ambitious than non-first-borns. Greek-affiliated students are more politically conservative than other students. Students with tattoos tend to be more ambitious than other students. Students who frequently hunt or fish are more patient than other students. Male students are more likely to make risky financial decisions than female students. Students who work a job while in school are more responsible than other students. 6 6

Google Forms Survey 7 7

Collecting Survey Responses Students have two choices for data collection: • In-person: Collect data in person at a variety of locations on campus by asking students to fill out their survey on their phone/tablet/computer. • Email: Send their survey link in an email to a random sample of students obtained using the on-line university directory. Response rates are typically 10%. The minimum sample size allowed is 30 but collecting more data is encouraged. 8 8

Google Forms Excel File with 3 Downloaded Responses. 9 9

Correlation Analysis Using Excel 10 10

F-test on Variances t-test on Means 11 11

My Experience with the Project Over the Past 5 Years • • Students seem to enjoy doing the project. Very detailed step-by-step instructions are provided. ALL students do very good in-class presentations. Written reports vary in quality, depending upon their understanding of the concepts and their procrastination habits. • I have to spend time re-writing their surveys (at the project proposal stage) and also grading written reports. However, I find it interesting and enjoy getting to know the students better. • Typically I have 2 classes of 56 students, so 56 groups in total. 12 12

How to Handle ~3 Projects for the New, Advanced Course • Should I start by doing 1 additional project (multiple linear regression) and then add the 3 rd project in a later semester ? • Does anyone have suggestions for resources concerning identifying a good MLR project ? • Should the instructor guidance be reduced in the 2 nd and 3 rd projects ? The reason I hesitate is that grading is easier and quality much better if I provide detailed instructions. • Audience suggestions and comments ? ? ? 13 13