ststs 4

Question 1

something missing with experiment?

Answer 1

blind? control? randomized unit selection and assignment of treatment?

Question 2

importance of random sampling vs voluntary/etc

Answer 2

reduces self-selection bias, favoritism. Results can be inferred to a larger population

Question 3

randomly assigning treatments

Answer 3

to create roughly equivalent treatment groups, allow inferences for cause and effect

Question 4

systematic random sampling: pick every ________ unit in the line

Answer 4

population size /sample size

Question 5

When samples larger

Answer 5

• margin of error smaller
• more precise
  -estimates closer to truth
• variability down

Question 6

What does it mean to NOT be statistically significant?

Answer 6

Difference between means/proportions of treatment X and treatment Y not large enough to rule out random chance as a result of random assignment as an explanation for the variation in results.

Question 7

What does it mean to be statistically significant?

Answer 7

Assuming that all happens by chance alone, differences in means (dots on dotplot) that occur less than 5% of the time by chance alone are considered stats. significant