Day 20: Statistics for Business Applications

Comments:

• Results were mixed - it was a disaster for some. (The distribution)
• Vocabulary hurt many: did I not tell you to know vocabulary?
  • interval estimators
  • unbiased estimator
  • consistent estimator
  • null and alternative hypotheses
  • p-value
• You didn't draw enough pictures.
• A key is available on-line.

• Proportions p and associated C.I.s (Section 12.4)
  • Test statistic for p (recall from 9.3)
  • Confidence interval estimator of p (when p is unknown!)
    • We can try to use p hat (the estimate of p)
    • Conservative standard error: use p=.5
    • Nielson example (recode, using Manip->code->numeric to numeric: 1 -> 1, 2:5 -> 0)
  • Sample size calculations for p