Industrial and Manufacturing Engineering 786 Applied Biostatistics in Ergonomics Spring 2012 Kurt Beschorner Note: This syllabus is not finalized and is subject to change up until the start of the class. Office: USR 201C Phone #: 229-6403 Email: beschorn@.uwm.edu Office Hours: W 4:30-5:30 p.m. and by appointment Class Data: T 5:30-7:20 p.m. (Lecture) R 5:30-7:20 p.m. (Laboratory) Catalog data: Textbooks: 490-786 Issues in Ergonomics: Measurement and Statistics. 3 cr. G. Statistical methods used in ergonomic studies to analyze, summarize, and report measurements and data. Jointly offered with & counts as repeat of NURS 786 & OCCUTHPY 786; with laboratory (2 HR LC & 2 HR LA/WK). Prereq Grad St., IND ENG 580(P); a course in statistics or cons. Instr. Dawson-Saunders, B. and Trapp, R.G., 1994, Basic & Clinical Biostatistics. Norwealk: Appleton & Lange. Portney, L.G. and Watkins, M.P., 2000, Foundations of Clinical Research: Applications to Practice, 2 nd Edition, New Jersey: Prentice-Hall. Handouts SPSS for Windows, Latest Ed., Student Version. References: Hennekens, C. H. and Buring, G.E., 1987, Epidemiology in Medicine. Boston: Little & Brown Co. Selected readings from current ergonomics literature. Goals: To emphasize the importance of statistics in ergonomics studies. To provide an understanding of of different types of epidemiological studies. To show students how to design and analyze ergonomics data using appropriate statistical methods. To show students how to determine sample size. Objectives: 1. Critique epidemiological studies in ergonomics. 2. Calculate inter- and intra subject variability and descriptive statistics. 3. Define and calculate incidence rate, prevalence, severity rate, odds ratio, relative risk, and rate ratio. 1
4. Estimate confidence intervals and test hypotheses. 5. Estimate and compare two or more means. 6. Estimate and compare proportions. 7. Fit linear, multiple and logistics regressions to experimental data. Teaching Methods: Lecture Laboratory Individual advisement 2
Applied Biostatistics in Ergonomics I & ME 786 (3 Credits) Instructor: Prerequisites: Kurt Beschorner, Ph.D. Grad. St.; I & ME 580; a course in statistics or cons. Instr. Week: Topics: Chapter:Pages 1. 1/24 Research Process Introduction to Epidemiology 2:7-22 Descriptive and analytical epidemiological studies Descriptive studies: Case report/case series Analytical studies Cross-sectional studies Case-control studies Retrospective cohort studies Prospective cohort studies Strength and limitations Critique of selected ergonomics studies 2. 1/31 Presenting and Summarizing of data 3:23-60 Scales of measurement Measures of central tendency Measures of spread Measures of nominal data Adjusted rates Relationship between two characteristics Incidence rate, Prevalence rate, Severity Rate Sensitivity and Specificity 3. 2/7 Probability 4:61-80 Independent and joint events Population & samples Methods of sampling (Random, systematic and stratified) Population parameters and sample statistics Binomial distribution Poisson distribution Normal distribution 3
4. 2/14 Single mean 4:80-90 Sampling distribution of the mean Center limit theorem Standard deviation v. standard error Confidence intervals & confidence limits Hypothesis testing 5. 2/21 Estimating and comparing single mean 5:93-109 t-distribution Confidence interval for the population mean Test of hypothesis for the mean Type I, Type II error 6. 2/28 Estimating and comparing proportions 5:110-131 Confidence interval Hypothesis test for a proportion Non-parametric procedure Sign test (Wilcoxon signed rank test) 7. 3/6 Review & Exam 8. 3/13 Comparing two Means 5:114-118 6:134-158 Matched or paired design (t-test for paired design) McNemar test for Proportions t-test for two independent means Chi-square test for two independent proportions Sign test Wilcoxon signed ranks test 9. 3/27 Comparing three or more means (ANOVA) 7:162-172 Factor Treatments Model Error or residual Fixed-effects model Random-effects model Assumptions 10. 4/3 Two-way ANOVA 7:175-186 Multiple comparison procedures Randomized factorial design Randomized block design Confounding Latin Square design Repeated measures design 4
Analysis of covariance 11. 4/10 Associations and Relationships 3:47-54 8:190-216 Pearson Product Moment Correlation coefficient t-test for correlation Coefficient of Determination Fisher s z transformation Confidence interval Spearman s rho Relative Risk Odds Ratio Confidence intervals for the relative risk and odds ratio 12. 4/17 Simple linear regression 8:202-216 Homogeneity Standard error of estimate Inferences Analysis of residuals Regression with repeat measures Common errors 13. 4/24 Multiple regression 8:212-216 Stepwise multiple regression Polynomial regression Missing observation 14. 5/1 Logistic regression 10:261-266 15. 5/8 Inter-rater Reliability 5:118-121 Kappa ICC 16. 5/15 Take-home final exam: due 5/19 at 12:00PM Grading System Midterm 20% Quizzes 20% Laboratory assignments 30% Take home final exam 30% 5