Lesson 8: Estimation

• Confidence interval:
  • Take a sample and get the mean
  • What is the probabilty that the mean weight of the population is within a range of the sample mean
• z-scores that bound 95% of a normal distribution
  • -1.96
  • 1.96

• 95% confidence interval for a sample mean
  • sample_mean - SE < mean < sample_mean + SE
• General idea:
  • The bigger the sample, the more confident one can be that mean derived from it matches the population mean
  • SE = population_standard_deviation / sqrt(sample_size) = a smaller value as the sample size gets large
• Critical values of z
  • 2.33 = 98%
  • 1.96 = 95%
• Practical use of Confidence Interval
  • Test a change (treatment) and collect data (dependent variables) from a population
  • Use it to determine where it lies on a sample mean distribition (z-score using SE)
  • Then, use z-table to calculate probability of getting that