# Instructor Class Description

Time Schedule:

Loveday L Conquest
Q SCI 482
Seattle Campus

### Statistical Inference in Applied Research I: Hypothesis Testing and Estimation for Ecologists and Resource Managers

Analysis of variance and covariance; chi square tests; nonparametric procedures multiple and curvilinear regression; experimental design and power of tests. Application to biological problems. Use of computer programs in standard statistical problems. Prerequisite: either STAT 311 or Q SCI 381. Offered: AW.

Class description

The objective of QSCI 482 is to develop the ability to translate a biological hypothesis into statistical terms for objective evaluation and hypothesis testing. We will use statistical decision rules with the following properties: objective evaluation, common understanding of the procedure (standard statistical conventions), and prior knowledge of the error rates for wrong conclusions associated with a given decision rule. We will also use a variety of statistical tests as a basis for discussion of issues of sample size and principles of good experimental design. Tests covered include one-sample, two-sample, paired-sample, and multi- sample comparisons (and their nonparametric equivalents), along with con- fidence intervals for estimated parameters where appropriate. Goodness-of- fit tests and chisquare tests for contingency table data are also included. WEEKLY homework assignments are required; there are 3 exams (the last exam occurs during finals week). We use SPSS statistical software (SPSS is also used in QSCI 483, the continuation course). ALL students in QSCI 482 should already be familiar with basic concepts of probability, estimation, and hypothesis testing as covered in a standard elementary statistics course like QSCI 381, STAT 311, STAT 220, or the equivalent. Basic concepts that students are expected to bring to the course include: some experience with hypothesis testing (meaning of Type I, Type II errors), use of basic probability distributions (e.g., normal, binomial distribution), and familiarity with certainly mathematical symbols and their usage (notably the summation sign). The ABILITY TO DO ALGEBRA also includes working with logarithms and exponents.

Student learning goals

Translate a biological hypothesis to be tested into an appropriate statistical hypothesis.

Choose the appropriate statistical method to apply, and to apply it correctly.

Relate statistical results back to the original biological question.

Understand standard statistical concepts such as null and alternative hypothesis, Type I and Type II error, power of a test, etc.

Be familiar with the standard ways of expressing parameters that describe a distribution.

General method of instruction

Classes will be largely lecture with occasional class activities when appropriate. For notetaking, students must print out topics in a timely manner from the course website.

Recommended preparation

Basic concepts that 482 students are expected to bring to the course include the following: some experience with hypothesis testing (including the meaning of Type I error, or level of significance, Type II error, use of basic probability distributions such as binomial, Poisson, normal). Also familiarity with mathematical symbols (notably the summation sign) and ABILITY TO DO ALGEBRA - this includes things like logarithms and exponents, and being able to solve equations (e.g., a single equation in a single unknown). HWs will be due on Fridays, with office hours (staffed by Prof. Conquest and the TAs) in MGH 291 during the week.