Steven Crane Compiled Section Handouts. Human Behavioral Biology 2012

Recognizing Relatives

Things to know for scientific findings:

• 
  What scientists mean by stating a degree of confidence about a finding.
  
• 
  The difference between the consistency of a difference and the magnitude of a difference, as played out on the population level.
  
• 
  Ditto on the individual level.
  

And know for Recognizing relatives

• 
  Why it makes sense to be able to recognize relatives.
  
• 
  How the MHC makes instinctual recognition of kin possible.
  
• 
  What sorts of sensory cues can be used for kin recognition.
  
• 
  Examples of cognitive derivation of kin in non-human species.
  
• 
  Non-cognitive modifiers of kin recognition in humans.
  

And know for Ethology

• 
  The contrasting intellectual traditions and styles of behaviorism versus ethology.
  
• 
  What a FAP is, examples, including human ones.
  
• 
  How ethologists approach the issue of adaptation, in contrast to the approaches of other disciplines we’ve heard
  

about.

• 
  How ethologists go about studying and identifying releasing stimuli.
  
• 
  Examples of releasing stimuli in different sensory modalities.
  
• 
  Some examples of neuroethological approaches.
  

What's the difference between statistical significance and functional significance/importance?

• 
  Stats is just about how consistent a finding is, how unlikely it is to be caused by chance. Is the difference observed "real" or accidental? Larger sample sizes make for greater statistical significance.
  
• 
  Actual importance of a finding, or effect size, is how much of a difference did you find? Does it actually make any difference in the real world for the way we think and make assumptions and behave?
  
• 
  Generally, the smaller the effect size is, the larger your sample size needs to be to prove that it's statistically significant.
  
• 
  There's also usually huge overlap and a lot of variability with big wide bell curves for most of these findings that are not very important.
  

• 
  So say you know ethnicity A is, on average, 1/4 inch taller than ethnicity B, on average.
  
• 
  Pick out any given person of ethnicity A and any given person of ethnicity B. Knowing only ethnicity, how likely are you to predict which is taller? The closer to a 50-50% chance, the less important the difference is.
  

• 
  Is a finding important?
  

• 
  How much of the variability in a population is explained by the finding? If not very much then it's not very important. Similar to heritability and variation.
  
• 
  If knowing what's true about groups gives you huge predictive power for individuals of those groups, then that's a very functionally significant (important) finding.
  

• 
  So this is the shortcoming of individual prediction based on overlapping group data.