WLD final review pt 2

• Statistically independent samples


  The value of each sample is not influenced by the value of other samples
• Taking a larger number of samples


  Decreases the standard error
• Probability of an event

  Can fall between 0 and 1
• Systematic sampling


  • Sampling on a grid; the one-in-k method
• Population index


  Counting bird vocalizations
• Point count


  To survey singing birds
• Optimal quadrat size


  Minimizes sample variation
• 95% confidence interval on an estimated x bar

  The interval which has a 95% chance of containing the true mean
• Stratified-random sample method for sample allocation

  • Uniform; optimal
• S^2


  Estimator for sigma^2
• Skew


  How symmetric a distribution is
• Methods to select a simple random sample


  • Draw numbers from a hat; use sample() in R; throw dice
• Power analysis


  To estimate the number of samples required to detect an effect of a given size
• 1. values


  The probability the observations occurred by random chance
• Standard error


  Equivalent to 68% confidence interval
• Autocorrelation


  Can only be positive
• Detectability


  The probability of an animal being detected given it is present
• Effect size


  The size of the difference between two populations
• Census


  A survey method where every individual in a population is counted
• Optimal quadrat shape


  • Elongated rectangles
• Variables that can be used to stratify a sample


  • Sex (male vs. female)
  • Age
  • Habitat
  • Population density
• Median


  The value of the sample where half the samples are large and half are smaller
• Type 1 error rate in hypothesis testing

  5%; Greek letter is alpha
• i.i.d. (independent identical distribution)


  Important for random variable draws in sampling design to be independent and identically distributed
• Parameters of standard normal distribution


  μ=0, σ^2=1