MBL/JBPC – Ten Lectures on Ecological Statistics


(whale shark – www.cancuncd.com)

Ten Lectures on Ecological Statistics

John Bunge, Department of Statistical Science, Cornell University


Marine Biological Laboratory
9:00-9:50 am M-F August 2-6, & 9:10-10:00 am M-F August 9-13, 2010

Dates & Locations:

Monday, August 2: Loeb G70
Tuesday, August 3: Loeb G70
Wednesday, August 4: Loeb G70
Thursday, August 5: Loeb G70
Friday, August 6: Speck Auditorium
Monday, August 9: MRC 210
Tuesday, August 10: MRC 210
Wednesday, August 11: MRC 210
Thursday, August 12: MRC 210
Friday, August 13: MRC 210

This series of talks will provide a self-contained introduction to statistics for biologists and ecologists, from basic concepts to current research. The course will enable the listener to understand statistical discourse in the applied research literature, to communicate with statistical consultants, and to carry out analyses with guidance from consultants or specialized software.

  • Week1
    • Fundamentals. Population and sample. Quantitative & qualitative variables.
      Probabilistic modeling of sampling processes. Parameters and statistics; statistical inference; parametric and nonparametric approaches. Estimation: estimates, standard errors,andconfidenceintervals. Hypothesis testing: type I & II error; power of tests. Simultaneous (multiple) inference and the Bonferroni correction.
    • Basic data analysis. Inference for a single variable. Relationship of two variables: qualitative-qualitative (contingency tables); quantitative-quantitative (regression); quantitative-qualitative (logistic regression & data mining); qualitative-quantitative (ANOVA and linear models).
  • Week2
    • Selected statistical methods in biological and ecological modeling. Analysis of biodiversity: estimating species richness and other diversity indices, using parametric and nonparametric techniques. Comparison of populations: Jaccard index and other measures. Variation of population characteristics as a function of biogeochemical predictors. More complex modeling of diversity vs. predictor variables: canonical correspondence analysis and alternative procedures.

Powerpoint Slides for this lecture series.

Print Friendly