Course Syllabus for PlSc 724 - FIELD DESIGN I Fall Semester - 2011 INSTRUCTOR: Rich Horsley Office 166C Loftsgard Hall or 109 Waldron Hall Office phone 231-8142 or 231-8924 Cell phone 793-1382 e-mail [email protected] web page: http://www.ndsu.nodak.edu/ndsu/horsley/ Course Description: PlSc 724 is a lecture course that discusses different statistical techniques for the analysis and interpretation of biological problems. Statistical techniques to be used include analysis of variance, simple linear regression, and simple correlation. Topics related to the planning of experiments to test hypotheses related to biological problems also are discussed. Prerequisite: An introductory course in statistics Required Text: Design and Analysis of Experiments, 6th Edition. 2005. D.C. Montgomery. Goals of PlSc 724: The broad goal for this course is to instruct students how to properly plan experiments, analyze data, and interpret results associated with testing hypotheses related to biological problems. Outcome 1 Students will be able to comprehend concepts needed to plan experiments to test hypotheses. These concepts include experimental error, replication and its function, relative precision, error control, and randomization. Outcome 2 Students will comprehend three experimental designs: completely random design, randomized complete block design, and Latin square design. For each design, students will know: the proper randomization procedure, how to describe the design, advantages and disadvantages, how to partition total degrees of freedom and sources of variation, the linear additive model, how to write expected mean squares, how to calculate estimates for missing data, how to do the analysis of variance, how to make tests of significance, and how to interpret results of significance. Outcome 3 Students will be able to choose the correct experimental design to test hypotheses related to biological problems. Outcome 4 Students will comprehend the use of simple linear regression to analyze and interpret results from experiments related to biological problems. Outcome 5 Students will comprehend the use of simple correlation to analyze and interpret results from experiments related to biological problems.Grading: Homework - ten homework assignments (15%) Two lecture examinations (25% each) Final exam - Comprehensive (35%) Thursday December 15, 3:15 to 5:15 PM This course is graded on a curve. The level of difficulty of the examinations and homework determines the gradelines for the curve. Yet, all scores of 90% or above are guaranteed an A, and scores of 80 to 89.9% are guaranteed a B. Statement on Academic Honesty Statement on Students with Disabilities or Special Needs Probable Dates When No Class Will be Held (Make-up lectures will be held) To be announced during class and on the course’s Web page. Any students with disabilities or other special needs, who need special accommodations in this course are invited to share these concerns or requests with the instructor as soon as possible. All work in this course must be completed in a manner consistent with NDSU University Senate Policy, Section 335: Code of Academic Responsibility and Conduct. http://www.ndsu.nodak.edu/policy/335.htmPlSc 724 TOPIC OUTLINE STATISTICAL REVIEW Types of variables Populations vs. Samples Three measures of central tendency Three measures of dispersion Variance of the mean and standard error Coefficient of variation Linear additive model PLANNING EXPERIMENTS Types of experiments Items to consider in planning experiments Experimental units Replication Choice of design Randomization HYPOTHESIS TESTING Type I error Type II error Power of the test Steps in testing hypotheses Testing the hypothesis that µ is a specified value (t-test and confidence interval) COMPARISONS INVOLVING TWO SAMPLE MEANS Two sample means with equal variance (t-test, confidence interval, and F-test) Two sample means with unequal variance (t-test) COMPLETELY RANDOM DESIGN ANOVA for any number of groups with equal replication ANOVA for any number of groups with unequal replication ANOVA with sampling Linear models for CRD experiments Assumptions underlying ANOVAMEAN COMPARISON TESTS Least Significant difference (lsd) Tukey’s test Testing effects suggested by the data Orthogonal linear contrasts RANDOMIZED COMPLETE BLOCK DESIGN ANOVA for any number of treatments ANOVA with sampling Linear models for RCBD experiments Experimental error in RCBD experiments LATIN SQUARE DESIGN ANOVA for single square ANOVA for repeated squares DIFFERENT ARRANGEMENTS USED IN EXPERIMENTAL DESIGNS Factorial Arrangements Split plot arrangements Split block arrangements Split-split plot arrangements COMBINED ANALYSIS OF EXPERIMENTS Combined analysis of experiments across locations Combined analysis of experiments across years Combined analysis of experiments across time and space REGRESSION AND CORRELATION Simple linear regression Simple