Topic outline

  • Painless biostatistics

    or

    BIOLOGICAL DATA ANALYSIS USING R

    a short and effective course to help you
    to perform by yourself the statistical analyses of your thesis.


    to be taught to PhD students at SLU-Alnarp,
    20-24 june 2005

    Earn 2 credits by completing the course!



  • Topic 1

    Biological hypotheses and their translation into statistical models

    After this class you will be able to
    • formulate scientific hypotheses with biological meaning
    • include such hypotheses into a cartesian plot, by identifying which are the (dependent and independent) variables involved and what are the mathematical relationship between such variables
       
  • Topic 2

    R system and Moodle basics

    After this class will you be able to
    • access the course's site and interact with Moodle System, changing your personal profile, reading text, posting questions and exercices, getting files
    • perform basic computations in R
    • plot simple graphs
    • read and include external data files into R enviroment
    • save your work
    • learn how to learn R-system
    Suggested reading:

    Crawley, chapter 2
  • Topic 3

    Probability, variance, and how to measure them.

    After this class you will be able to:
    • Measure the variability of your data
    Anova

    After this class you will be able to:
    • Measure the degree of uncertainty when stating that two sets of data belong to the same probability distribution
  • Topic 4

      Nested and Split Plot anovas

    After this class you will be able to:
    • Avoid running into pseudoreplication when analysing a data set composed by hierarchicaly structured experiments, whose variables come from different spatial scales.
  • Topic 5

    Regression models

    After this class you will be able to:
    • analyse models whose independent variable(s) is(are) quantitative (that is, x-var is continuous)
  • Topic 6

    Ancova models

    After this class you will be able to:
    • Analyse models composed by a quantitative and a qualitativa variable
    • Interpret the biological meaning of a statistical interaction
  • Topic 7

    Statistical modelling: building, criticising and simplifying models

    After this class you will be able to:
    • translate biological questions into statistical models
    • build different statistical models for the same set of variables
    • check whether the model achieved is coerent to its underlying statistical assumptions
    • propose alternative, simpler and statistically sound models which could explain the same data set
    Crawley, chap 13
  • Topic 8

    Dealing with non-normal data

    After this class you will be able to:
    • perform a statistical analysis of a non-normal data set, without having to force normality via y-var transformations
  • Topic 9

    Count data models

    After this class you will be able to:
    • Analyse data whose response variable is in the form of counts (ex. number of species, number of times something happened)
  • Topic 10

    Proportion and binary data

    After this class you will be able to:
    • Analyse data whose response variable is a proportion (ex. percent of insects choosing left arm of a y-olfactomer)
    • Analyse data whose response variable binary (ex. effect of predation on presence/absence of prey)
  • Topic 12

    Survival analysis

    After this class you will be able to:
    • Analyse data on deaths or failures (time to death of insects submitted to insecticide, time to final decay of a pheromone dosis)
  • Topic 13

    Experimental design

    After this class you will be able to

    • design an experiment assurring true replication and randomization
    • spot pseudoreplication
    Suggested reading:

    Crawley, chapter 3
    Hurlbert