18.1 General

  • Any statistical model is simply a mathematical equation (or several)
  • We can plug in values (e.g. values for our explanatory variables) into this equation
  • The equation contains parameters (e.g. \(\beta\)), which we don’t know but we can estimate them
  • Normally, we try to estimate the parameters in our model (\(\beta\)s) so that the result, i.e. the values predicted by our equation (\(\hat{y}_{i}\)) are as close as possible to the observed values (\(y_{i}\)) of our outcome variable
  • There are various approaches for estimating parameters
    • OLS
    • Maximum Likelihood
    • Bayesian estimation (see Rstan)