Statistics Project

You are to choose a topic for your project.  It may be something related to your job or a special interest of yours.  You need to do the following things:

1. Choose a topic that you would like to use for your project.
2. Develop a hypothesis for your topic. This is some conjecture you would like to test.  You must use two continuous variables and a cause and effect
3. Collect data for your analysis. You need to have at least 50 sets of data.  You will also need to detail how you collected the data. If you intend to use a survey, it must be approved before you administer it.
4. Perform the appropriate statistical analysis on the data. This will include standard measures such as mean, median, mode, standard deviation, variance, correlation coefficients and regression lines.  You will need to detail how you examined the data, i.e. grouping, sampling, etc.

5. Create the appropriate diagram(s). This would include histograms, special distributions, etc. You must include a scatter diagram and fitted line plot.
6. Test your hypothesis. This would include indication of the type of distribution, linear regression testing, confidence level of your conclusion, basic statistics, etc.
7. Write up your results. You need to submit the final project that includes all documentation, diagrams, computer printouts, data, survey, etc. You need to state why you chose the tests or analysis that you used.  It is important that state your hypothesis and conclusion in the final project.

POINT VALUE: TOTAL                 100 points

Hypothesis                                          10

Data                                                    15

Appropriate Diagrams                     25

Statistical Analysis                             25

Written Summary                             25

An Example of a Topic for the project

This Report will examine the hypothesis that there is a positive correlation between Sat Verbal score and High School GPA.  Data was collected from the Focus database that accompanies the textbook for this course.  The analysis will include charts and diagrams, regression analysis and formal significance testing.

The data was collected from Arizona State students.  Five hundred observations of data were collected, gathering seven different variables for each observation.  Those variables were sex, high school GPA, Sat math score, cumulative GPA, SAT verbal score, age and total hours earned.

A scatter diagram on SAT verbal score and high school GPA indicates that there may be a correlation, but probably not a very strong one.  The regression analysis gives a regression equation of HSPGA = 2.30485 + 0.00158 SATV.  The coefficient of determination, r²  = 7.0 %.  This indicates that only 7 percent of the data can explained by the regression line.  This is not a very good percentage.

The correlation coefficient is .264.  This value is relatively close to zero, which indicates that there really is not a correlation between SAT verbal score and high school GPA.  At the a  = .05 significance level, the test correlation coefficient is .31 and at the a  = .01 the test correlation is .42.  Since there were 500 observations examined.  The results are significant.

A histogram on the high school GPA indicated that most of the students had a high school GPA of 3.0 or higher.  The mean for high school GPA was 3.0271 and the median was 3.0.  The standard deviation on high school GPA was .4981.

For SAT verbal score the median was 450 and the mean was 455.88.   The standard deviation was 82.93.  The histogram on SAT verbal indicated more of the students were in the range of 400 – 450, which is not very good.

The analysis done in this report indicates that there is no correlation between high school GPA and SAT verbal score.  The linear regression indicated that only 7 % of the data is be explained by the regression line and 93% of that data is unexplained.   The scatter diagram shows no real correlation.  From this analysis, the relationship between high school GPA and SAT verbal score cannot be determined.  Analysis of correlation on high school GPA predicting SAT Verbal score should be done.