|
QMM 241 – Summer 2012 |
Mini-Projects |
Prof. Doane |
|
Instructions Each mini-project may be done partners (submit only one
report). You will be graded on how well you use the tools from the textbook
and lectures, and on the effectiveness of your written report (e.g.,
spelling, grammar, style, organization, tables, graphs). For guidance on
effective writing, see Appendix I in the textbook. For calculations and
graphs, use MegaStat or Minitab or Excel (or any convenient combination of
these tools). MP-4 and MP-5 have more weight. |
MP-1 (a) Go to http://www.sba.oakland.edu/faculty/doane/downloads.htm and download the spreadsheet on coffee temperature for Bruegger’s and Panera. (b) Perform a t-test for two independent samples using both Case 2 and Case 3. What is your conclusion? (c) Perform an F-test for equal variances. What is your conclusion? (d) Make a histogram and normal probability plot for each sample. Are the data normal? Why is this important? |
|
MP-2 Choose one single-factor ANOVA data set from exercises 11.2, 11.4, 11.22, 11.26, or 11.28. (a) State your hypotheses, degrees of freedom, critical value, and decision rule for each statistical test that you perform. (b) Interpret the visual display of means (if MegaStat) or confidence intervals (if Minitab). (c) Use MegaStat or Minitab to perform the one-factor ANOVA. (d) Draw appropriate conclusions from the F-test. (e) Which pairs of means (if any) differ, according to the Tukey test? Explain. (f) Use Hartley’s Fmax test and/or Levene’s test to check for equal variances. Discuss the results.. |
MP-3 (a) Do exercises 17.49, 17.50, and 17.51. Use Minitab or MegaStat for the control charts and control limits in 17.50. Write a concise report containing cut-and-paste relevant charts and a summary of your findings. |
MP-4 The instructor will assign you a team and a company. Download the spreadsheet with quarterly revenue data from http://www.sba.oakland.edu/faculty/doane/downloads.htm. (a) Briefly describe the company’s history (e.g., http://www.mergentonline.com/basicsearch.php or http://www.finance.yahoo.com), its products and services, competition, and market conditions (supply, demand, substitutes). (b) Fit several trends (e.g., linear, quadratic, exponential) using Excel or Minitab or MegaStat. (c) Interpret each fitted trend equation. (e) Use Minitab or MegaStat to calculate seasonal factors. Is there seasonality? (f) Forecast the next 4 quarters, both with and without seasonality. Bottom Line: Which trend model (if any) would yield credible forecasts? If none, then what? |
MP-5 The instructor will assign you a team and a dependent variable from one of the state or nations databases. (a) Propose a reasonable model of the form Y = f(X1, X2, ... , Xk). (b) Use regression to estimate the model. (c) Delete useless predictors until you feel that you have a parsimonious model, based on the t-values, p-values, standard error, and R2adj. (d) For the preferred model only, identify observations with high leverage and/or unusual residuals. (e) Check the residuals for evidence of non-normality, autocorrelation, and heteroscedasticity (e.g., plot the residuals against Yfitted).Describe your findings. (h) Analyze the correlation matrix (collinearity) and VIFs (multicollinearity). Is variance inflation a serious problem (Klein’s rule)? Bottom Line: Was your regression helpful in understanding variation in Y? If you had more time, what would you do? Why did regression come last in this course? |