SCHOOL OF BIOLOGICAL & CONSERVATION SCIENCES

PIETERMARITZBURG CAMPUS 2008

BIOL701 Biology/Ecology Tools and Skills

Refresher Course on Statistical Analysis and Experimental Design Course Outline & Content

Faciltated by: Craig Morris (room 107b, B-block; [email protected]) and Caryn Penderis (room 107a, A-block; [email protected]). Course schedule: See Honours Techniques 2008 Timetable Exptl Design and Intro Stats. Two sessions per week for class discussions (in JB Scott room) and practicals (in School LAN) from 6 February 12 March 2008. The theory test will take place on 12 March (8h30 10h30) and a computer practical test on 10 March (14h00 - 17h00). Aim: To provide you with skills for planning and implementing scientific experiments / trials and for analyzing and interpreting biological and ecological research data, including the ability to critically interpret and evaluate statistical results presented in research reports and published scientific papers. Mode of delivery: This refresher course is designed to remind you of the basic statistical techniques covered in your second year of study (Biol200) and to deepen and extend your knowledge on these and other statistical methods. The course largely comprises self-study through extensive reading around some essential topics regarding the design and layout of experiments and statistical analysis of quantitative univariate data of various kinds to answer particular research questions. Though practical application of the methods (via computer packages) will be demonstrated, the emphasis of this course is on you gaining (and demonstrating) an understanding of the key statistical concepts underlying the choice and application of appropriate statistical test(s) and appropriate interpretation of and inference from the results. During the classroom sessions, one member of the class will be randomly selected to give a 10-minute presentation (PowerPoint is not mandatory) on one of the designated topics, followed by group discussion of issues around that topic. Each student will get at least two chances to present during the course (see assessment). During prac sessions, students will analyse and interpret various data sets, using either SPSS or Genstat, as practice for the practical test (see assessment). Assessment: Please note that there are no examinations for this component, i.e. the class record comprises 100% of the final mark for this section, which counts eight credits (of a total of 16C for Biol701). The class mark will comprise marks from one theory test (50% externalled), a practical test (30%) and class presentations (20%). Course materials, readings and data sets will be made available on the OLS at: http://ols.ukzn.ac.za/system/interface/interface.php

COURSE CONTENT TOPICS FOR DISCUSSION (See attached timetable for expected date of presentation & discussion) Fundamental concepts of statistics 1) Why bother analyse data that is obviously so different? 2) What is the difference between descriptive & inferential statistics? 3) What are some useful measures of central tendency and variability? 4) What are some useful measures of precision? 5) Why the obsession with the normal distribution? 6) What are Type I and Type II errors? 7) Why test for no difference (null-hypothesis testing)? 8) Are non-parametric statistics the answer to a distribution-free life? Treatment and experimental design and layout 9) What is an experimental unit? 10) Why is replication required and how should it be applied? 11) What is the role of random selection/allocation in experiments? 12) When are unstructured and structured treatments useful? 13) What is the value of a factorial design? 14) What designs can be used to cater for variation in space and time? Two-group comparisons and one-way ANOVA 15) When should an independent or paired t-test be used? 16) What is the purpose of analysis of variance (ANOVA)? 17) What is a sum of squares and an F-test? 18) What are the joys and pains of using post-hoc tests? 19) Why worry about the assumptions of ANOVA? Multi-way ANOVA, ANCOVA and repeated-measures analysis 20) What are treatment main effects and interactions? 21) How are interactions interpreted and represented graphically? 22) What is the purpose a covariate in analysis of covariance (ANCOVA)? 23) When is repeated-measures ANOVA required? Correlation and regression for understanding and modelling relationships 24) How are correlation and simple linear regression related? 25) What statistical tests are required in regression 26) When are polynomial and non-linear regression models useful? 27) What is the value of multiple regression? 28) How are coefficients in multiple regression interpreted? 29) How are potential problems in regression models diagnosed? 30) Are there alternatives to stepwise regression for identifying the best model? Additional tests and methodology 31) What is the chi-square test and when is it used? 32) What methods can be used for binomial data? 33) What is the Generalized Linear Model? 34) Why is the information-theoretic approach different to null hypothesis testing?

TIMETABLE FOR PRESENTATION OF TOPICS AND COMPUTER PRACTICALS Day Wed Feb 6 Fri 8 Tue 12 Thu 14 8h30 10h30 Introduction & course outline Introduction to statistics Topics 1-4 Introduction to statistical testing Topics 5-8 Choice & allocation of experimental units Topics 9-11 Treatment structure & Experimental layout Topics 12-14 Prac: Randomisation of experimental units Two-group tests & one-way ANOVA Topics 15-19 Prac: Two-group tests & one-way ANOVA Multi-way ANOVA & ANCOVA Topics 20-23 Simple, polynomial & non-linear regression Topics 24-26 Multiple regression Topics 27-30 Binomial & count data Topics 31-32 Stats in the 21st century Topics 33-34 Theory test Prac: Multi-way ANOVA & ANCOVA Prac: Simple, polynomial & nonlinear regression Prac: Multiple regression Prac: Binomial data & Practice for prac test Prac test Prac: Free Prac: Descriptive & inferential statistics Prac: Reporting & graphing results Prac: Literature survey of useful experimental designs in biology and ecology 14h00 17h00

Mon 18

Tue 19 Wed 20 Thu 21 Mon 25 Wed 27

Mon Mar 3 Wed 5 Mon 10 Wed 12

Please note: Date of presentation of topics and practical schedule might be subject to change at short notice.

LIST OF REFERENCE MATERIAL TO READ 1) Biol200 notes and updated manual (2008) 2) Reliable web resources, such as: StatSoft Electronic Textbook at: http://www.statsoft.com/textbook/stathome.html 3) Books in the UKZN library, some of which are listed below: Around 519.5 on the shelf
Beyond ANOVA : basics of applied statistics. Miller, Rupert G. 519.5352 MIL Design and analysis of experiments. Hinkelmann, Klaus. 519.62 HIN V.1 Distribution-free tests. Neave, Henry Robert. X 519.5 NEA Handbook of parametric and nonparametric statistical procedures. Sheskin, David J. X R 519.5 SHE How to think about statistics. Phillips, John L. X 519.5 PHI Modern statistics for the life sciences. Grafen, Alan. X 519.5 GRA Nonparametric statistical inference. Gibbons, Jean Dickinson. 519.5 GIB Practical statistics for environmental and biological scientists. Townend, John. X 519.5 TOW Principles and procedures of statistics : a biometrical approach. Steel, Robert G. D. X 519.5 STE Statistical methods. Snedecor, George Waddel. 519.5 SNE Statistical modelling using genstat. McConway, Kevin J. X 519.5 MACC Statistics : a guide to the unknown. Peck, Roxy. 519 PEC Statistics : a guide to the unknown. Tanur, Judith M. X 519.5 TAN Statistics : a spectator sport. Jaeger, Richard M. 519.01 JAE Statistics for people who (think they) hate statistics. Salkind, Neil J. 519.5 SAL Statistics for research. Dowdy, Shirley. X 519.5 DOW Statistics for the environment. Barnett, Vic. 333.70724 BAR Statistics in the 21st century. Raftery, Adrian E. 519.5 RAF The cartoon guide to statistics. Gonick, Larry. X 519.5 GON The visual display of quantitative information. Tufte, Edward R. 519.5 TUF

Around 574.01 574.0725 on the shelf

An introduction to experimental design and statistics for biology. Heath, David. X 574.015195 HEA Biological data analysis : a practical approach. Fry, John C. X 574.0725 FRY Biometry : the principles and practice of statistics in biological research. Sokal, R. R. X 574.0725 SOK Biostatistical analysis. Zar, Jerrold H. X 574.0725 ZAR Choosing and using statistics : a biologist's guide. Dytham, Calvin. X 574.0725 DYT Encyclopedia of biostatistics. Armitage, Peter. X R 574.015195 ARM Experimental design and data analysis for biologists. Quinn, Gerry P. X 574.0725 QUI Introduction to biostatistics. Robert R. Sokal and F. James Rohlf. X 574.0725 SOK Introductory statistics for biology students. Watt, T. A. X 574.0725 WAT Practical statistics for field biology. Fowler, Jim. 574.0725 FOW Statistical methods in biology. Bailey, Norman Thomas John. X 574.0725 BAI Statistics for biologists. Campbell, Richard Colin. X 574.0725 CAM

Around 630.724 630.725 on the shelf

Agricultural field experiments : design and analysis. Petersen, Roger G. X 630.725 PET Design of experiments for agriculture and the natural sciences. Hoshmand, A. Reza. 630.724 HOS Statistical experiment design and interpretation. Collins, C. A. X 630.724 COL Statistical methods in agriculture and experimental biology. Mead, Roger. X 630.725 MEA