Final Exam Flashcards
Null hypothesis
statistical hypothesis that one variable has no relation to another; THIS is the hypothesis that is actually being tested in science
Heuristic
done just for the sake of increasing your knowledge base
Observational study
collect data without influencing the subject(s)
Experimental study
have control and experimental settings; test for differences; should ideally be double-blind (i.e. neither subject(s) or researcher(s) know which group is control and which is experimental)
Operational definition
defines how you are going to measure what you’re testing
Nominal measurement
No numeric meaning; qualitative categories within which individuals are counted (i.e. sex, species, ethnicity); chi-squared test
Ordinal measurement
Data can be organized by rank order, but relative degree of difference between ranks can’t be calculated (i.e. dominance rank, conservation status, health, guilt); Spearman’s correlation, Mann-Whitney U
Interval measurement
Allows for degrees of difference between data points, but not a ratio between them (i.e. temperature - 20degC is not “twice as hot” as 10degC); Pearson’s correlation, Student’s t
Ratio measurement
Have a zero scale and can be expressed in ratios (i.e. mass, length, distance); Pearson’s, Student’s t
Pearson’s correlation (r)
equires linear relationship for interval data; PARAMETRIC
Spearman’s correlation (rs)
used with non-linear, ordinal data; NON-PARAMETRIC
P Value
Assuming null hypothesis is true, what is the chance of observing statistics at least as extreme as ones actually observed if there really is no correlation between X and Y?
Alpha Level: cut-off point above which the null hypothesis is accepted; usually > 0.05
Coefficient of determination
r^2
fraction of variance that is shared between X and Y
Student’s t-test
For PARAMETRIC data
By W.S. Gossett
Assesses whether the means of two groups are statistically different from one another
Requires three things: [1] variables must be normally distributed within each group; [2] standard deviations must be similar; [3] dependent variable must be interval or ratio
Mann-Whitney U test
For NON-PARAMETRIC data
Does not make a distribution assumption, so is not as powerful as a t-test on normally-distributed data
Dependent variable can be ordinal, interval, or ratio
Chi-squared test
For NON-PARAMETRIC data
Compares observed counts of a variable to expected counts
Exaptation
a trait that performs a current adaptive function but arose either for another function or with no function at all; includes preadaptations and spandrels
Preadaptation
when a feature is secondarily co-opted for a function other than the one for which it originally evolved
Phylogeny
evolutionary relationships within/between taxonomic levels; esp. patterns of lines of descent
Phylogenetics
taxonomical classification of organisms based on their degree of evolutionary relatedness
Polytomy
unresolved internal node of a cladogram that has more than two immediate descendents
Cladogenesis
Speciation - multiple new lineages can originate from a single ancestor in a short geologic period
- Should be bifurcating
Homoplasy
Convergent character - shared between species but was not present in their common ancestor (i.e. wings in birds and bats)
Homology
characters shared between species that was also present in their common ancestor
Symplesiomorphy
trait that is found in the ancestor of a group but only some of its descendants; do not indicate phylogenetic groups; shared ancestral homologies
Synapomorphy
a trait that is unique to a group of species and their ancestor; indicate phylogenetic groups; shared derived homologies; in a new state from the ancestors of a group and their common ancestor