**Course Description**: This course is a second course in applied
statistics. Statistics involves numbers in context. The numbers
you analyze are numbers of something and the results should be described
in terms of that thing. Thus, statistics is its own peculiar
mixture of common sense, mathematics, and communication skills.
This course emphasizes all of these aspects.

The topics covered include simple and multiple regression, analysis of variance with fixed and random effects, multiple comparison procedures, model selection and verification procedures, categorical data analysis, odds ratios, and prediction of binary responses.

**Assignments**: The sole basis for your grade in this course
is your performance on the homework assignments. Each student's lowest 2 assignment grades will
be dropped from the final average - this should cover reasonable illnesses and excused absences. Most of the
homework problems involve analyzing real datasets and
writing brief reports summarizing your results.
These analyses and reports should
be taken seriously, done properly, and written well. When asked to
summarize your findings on a homework problem, you should write
a consultant's report that any lay person could understand, keeping
statistical jargon to a minimum.
Example acceptable write-up of an exercise. Assignments should be typed into
a word processor with graphics added as appropriate.

Students in the past have commented that they do not understand what I expect from these assignments and how I assign grades to their answers. These comments reflect the difficulty I have in assigning grades. Your write-ups need to be clear, grammatical, concise, and understandable (in English not in statistical jargon). Your statistics need to be appropriate to the problem. Some techniques may be better than others, some techniques may not be applicable, but, often, there is more than one good technique to address any given problem. Example: If your dataset consists of all 0's, 1's, and 2's, techniques that rely on rankings are unlikely to be applicable and you should be suspicious of using continuous-value based normal theory statistics, such as t-tests, but categorical and bootstrapping techniques may work well. Finally, you need to use common sense. Example: If the dataset contains a 100 lb cicada, you should figure that out, report it as a data entry error, and remove it from the data set (or correct it) before conducting your analysis. ALL of these aspects are critical to good applied statistics and should be taken into account when assigning grades.

Another comment that students have made is that the text is not an appropriate reference text for their research. This complaint is valid. The text and the software I choose to use are chosen for teaching purposes. The statistics text you use as a reference for your research and the software you need to use for your research may well vary from that used in class.

** Software:** We will use Minitab in class. You may use other software, but my
demonstrations and instructions will be for Minitab. Minitab is simple, menu driven but with
easy syntax for writing code when necessary, and cheap. You may purchase a year's license
through the IU's Math Stat center for $20
or a 5 month liscence directly through
the Minitab corporation
for $29.99. It is a wonderful program and you might consider buying a permanent copy for only $100.

## Tentative Syllabus | ||
---|---|---|

Date | Topics Covered | Homework Assignments |

Monday, August 28 | Ch. 1: Inferences and p-values | Ch. 1: 17 (no essay required); Ch. 2: 22, 23
Corrected Data Set
due September 6 |

Wednesday, August 30 | Ch. 1: Code for bootstrapping a p-value | |

Friday, September 1 | Ch. 2: Review of t-tests | |

Monday, September 4 | Ch. 3: Review and assumptions of t-tests | Ch. 3: 33 Additional (non-essay) problem: form a display like 3.5 in your text using Welch's t-test (without pooling variances) due September 13 |

Wednesday, September 6 | Ch. 4: Log Transformations | |

Friday, September 8 | Ch. 4: Alternatives to t-tests - nonparametric approach | |

Monday, September 11 | Ch. 4: Alternatives to t-tests - bootstrapping | Ch. 4: 31, 32
due September 20 |

Wednesday, September 13 | Ch. 4: Paired tests | |

Friday, September 15 | Ch. 4: Guide to t-tests | |

Monday, September 18 | Ch. 5: ANOVA - The Theory | Ch. 5: 21 (no essay required), 23
due September 27 |

Wednesday, September 20 | Ch. 5: ANOVA -In Practice | |

Friday, September 22 | Ch. 5: NonParametric ANOVA | |

Monday, September 25 | Ch. 5-6: Linear Contrasts and Random Effects |
Ch. 6: 22
due October 4 |

Wednesday, September 27 | Ch. 6: Multiple Comparisons | |

Friday, September 29 | Ch. 7: Simple Linear Regression | |

Monday, October 2 | Ch. 7: Simple Linear Regression |
Ch. 7: 26 (no essay required), 29
due October 11 |

Wednesday, October 4 | Ch. 7/8: Regression Diagnostics | |

Friday, October 6 | Ch. 8: Regression Diagnostics | |

Monday, October 9 | Ch. 8: Regression Diagnostics |
Ch. 8: 20
Corrected Data Set ,
24 (provide a truly useful plot or chart - minimize the effort required on he physician's part)
due October 18 |

Wednesday, October 11 | Ch. 9: Multiple Regression | |

Friday, October 13 | Ch. 9: Multiple Regression | |

Monday, October 16 | Ch. 10: Multiple Regression | Ch. 9: 20; Ch 10: 29
due October 25 |

Wednesday, October 18 | Ch. 10: Multiple Regression | |

Friday, October 20 | Ch. 10: Multiple Regression | |

Monday, October 23 | Ch. 11: Model checking | Ch 11: 24 (find the common names of the species and identify obvious data-entry errors); Ch 12: 21
due November 1 |

Wednesday, October 25 | Ch. 12: Variable Selection | |

Friday, October 27 | Ch. 12: Variable Selection | |

Monday, October 30 | Ch. 13: Two-Way ANOVA |
Ch. 13: 19
due November 8 |

Wednesday, November 1 | Ch. 13: Two-Way ANOVA | |

Friday, November 3 | Ch. 13: Two-Way ANOVA | |

Monday, November 6 | Ch. 13: Two-Way ANOVA | Ch. 14: 14, 15, 16 (these are related projects and can be written up into one extended essay) due November 15 |

Wednesday, November 8 | Ch. 14: ANOVA without replication | |

Friday, November 10 | Extra: Random Effects and Experimental Design | |

Monday, November 13 | Ch. 18: Proportions and Odds | Ch. 18: 18; Ch. 19: 17 due December 1 |

Wednesday, November 15 | Ch. 18: Proportions and Odds | |

Friday, November 17 | Ch. 19: Chi-Square, Fisher's Exact Test, Mantel-Haenszel | |

Monday, November 20 | Ch. 19: Chi-Square, Fisher's Exact Test, Mantel-Haenszel | See above homework |

Wednesday, November 22 | Thanksgiving Recess | |

Friday, November 24 | Thanksgiving Recess | |

Monday, November 27 | Ch. 20: Logistic Regression | Ch. 20: 15, 18 due December 8 |

Wednesday, November 29 | Ch. 20: Logistic Regression | |

Friday, December 1 | Ch. 21: Logistic Regression | |

Monday, December 4 | Ch. 21: Logistic Regression |
Ch. 21: 19 due December 13 |

Wednesday, December 6 | Ch. 21: Logistic Regression | |

Friday, December 8 | Review |

This page updated on August 25, 2006 by Elizabeth Housworth

Department of Mathematics

Department of Biology

Indiana University

Bloomington, IN 47405

Phone: 812-855-1960

Email: < ehouswor@indiana.edu >