BIOL3045 Lecture Notes - Lecture 3: Frequentist Inference, Reaction Norm, Reaction Rate
3 Jul 2018
School
Department
Course
Professor
Animal Ecological Physiology BIOL3045 460381099
Lecture 3: Experimental design.
Generalist-specialist trade-off
If phenotypes are heritable, then specialization to that environment occurs.
Variable environments favour performance of generalists.
The trade-off is: maximum performance of generalists is lower than that of specialists, their reaction
norm is wider but the peak is not as tall as specialists.
Similar to reaction rate vs. temperature where: specialists have small changes in temp with large
changes in reaction rates; generalists have small change in temp with small change in reaction rate.
Reading: Drummond, G. B. and Vowler, S. L. 2012. Different tests for a difference: how do we do
research? J. Physiol 590.2, 235-238.
• What is “frequentist” statistics? Statistics that determines how frequently would a result like this
be observed. Calculates false discovery rates using confidence intervals, effect size, power
analysis.
• What is the “straw man” in frequentist statistics? Set-up experiments to test for an effect, but
the logic behind the tests is to look for lack of support. Ho will never be true, because there will
always be some differences.
• What is the Bayesian logic? Bayesian approaches use experimental data to update this prior
knowledge (from experiments and theory).
• What are the strengths of Bayesian logic? Use of data, not the underlying distributions.
• What are permutation tests? Randomization of data to test how frequently one would get a
result as extreme or more extreme as the experimental results, relative to the total number of
permutations. Advantageous because most studies have relatively small sample sizes
• that cannot represent assumed distributions, also has no assumptions.
Document Summary
If phenotypes are heritable, then specialization to that environment occurs. The trade-off is: maximum performance of generalists is lower than that of specialists, their reaction norm is wider but the peak is not as tall as specialists. Similar to reaction rate vs. temperature where: specialists have small changes in temp with large changes in reaction rates; generalists have small change in temp with small change in reaction rate. Reading: drummond, g. b. and vowler, s. l. 2012. Statistics that determines how frequently would a result like this be observed. Set-up experiments to test for an effect, but the logic behind the tests is to look for lack of support. Randomization of data to test how frequently one would get a result as extreme or more extreme as the experimental results, relative to the total number of permutations.
