Statistical Analysis Project: Data Collection and Reporting

Topic Selection and Objectives

Task:

The purpose of this project is to encourage each student to apply statistical techniques to real-world phenomena. The project will entail data collection and statistical analysis using the theoretical tools developed in this course. The class project will culminate in a final report which includes an introduction to the statistical problem studied and concluding remarks and recommendations that can be made from the statistical analysis.

One of the objectives in this course is to give you an opportunity to sharpen your written and oral communication skills. You will create an effective PowerPoint presentation, record your voice narration, and upload your video file to the Moodle class website before the due date.

A. Topic

1. Your project can be doomed to failure by selecting an inappropriate topic. For example, you often select a problem scope or objective that is much too large for them to actually accomplish in a reasonable amount of time. Such projects are often referred to sarcastically as trying to "solve world hunger" or "boil the ocean".

2. When choosing a problem to study, be creative. Some of the past projects include:

 Product prices at two supermarkets or on-line book stores
 Service time at drive-through restaurants, grocery store express lanes, or post offices
 Drinking habits between Greek vs. non-Greek
 Average tips earned by waiters and waitresses
 Test scores in private and public schools
 Opinions between Democrats and Republicans
 Winning percentage between home games and road games

B . Type of Project

1. For the statistical project, you will collect sample data from two populations, and test for the difference between two population means, or population proportions, as we did in Session 12. Two-Sample Tests. Thus, your project belongs to one of the following three cases:

? Numerical data

Case #1. Comparing two independent samples (i.e., independent sample case)

Case #2. Comparing two related samples (i.e., paired or matched sample case)

? Categorical data

Case #3. Testing differences in two proportions (i.e., population proportions)

2. Be sure to finish Session 12 in the lecture notes and check the type of your class project before you start to collect sample data.

CC. Data Collection. Data Collection

1. Collect a data set from two populations relating to your work or business interests, with a sample size of n > 30 from each population (the more, the better!). The variable of interest could be either a quantitative (numerical) or a qualitative (categorical) variable.

2. To obtain the necessary data, you may (i) conduct a survey, (ii) make observations through an observational study, or (iii) design an experiment. In some cases, you may need to get permission to gather data. For example, if you want to watch a service facility such as a bank or post office, you should ask permission of the management of the facility. Be sure to inform them that you are using this only as a student project and in no way will the performance (good or bad) of their servers be publicized.
DD. . WWritten Reportritten Report

1. Perform a statistical analysis using the sample data set. You should use Microsoft Excel (or a similar computer program) to do all of the calculations and plotting. You should spend much of your time on the project interpreting the output from the computer and determining if any other analyses are necessary.

2. Write up the results as a report to upper-level management, either as a background summary report or as a proposal for action. Your report should include 6~10 pages (1.5 line spaced) plus an appendix, and should be based on the following format:

? Cover page that includes the title of the project and your name.

Describe the background and questions of interest clearly as if to an intelligent person who knows nothing about the details of the situation. Also, define the target populations of your interest, explain the sampling method, and justify its randomness.

? Statistical analysis (may be subdivided into several sections)

Analyze the data, presenting displays and results, explaining as you go along. Consider including some of the following:

(a) Explore the sample data using histograms, contingency tables, pie charts, or box plots for each population.
(b) Compute the sample statistics such as the sample means and sample variances, or the sample proportions.
(c) Compute appropriate estimates for population parameters and their standard errors.
(d) Find a confidence interval for each population parameter and for their difference.
(e) Describe the null and alternate hypotheses.
(f) Test the hypothesis and find the p-value.
(g) Interpret and explain your results