University:
Massachusetts Institute of Technology
Course:
15.501/516 | Accounting
Academic year:

2023
Views:

326

Pages:

11
Author:

Gabriel Anthony

Hypotheses testing I Different things are different ! Why test them? Can’t we just show mean difference? No, because of randomness / chance • Can we test that things are different? Logically it is very hard -- because they are • Variance Within vs. between subject designs Which is better? Why What are the advantages of each? What are the disadvantages of each? Hypotheses testing II H0 (the null hypothesis) H1(the real hypothesis) H0- No difference Group 1 = Group 2 [=, ≤, ≥] H1-Difference Group 1 ≠ Group 2 [≠, >, <] • • Hypotheses testing III H0 & H1-- Examples: Is a coin fair Gender and grades Healing with a new medication Ability to cheat Marriage over time For each please write H0 & H1 Hypothesis testing IV Why do we test a hypothesis we don’t believe in? What does it mean to reject H0? If H0 is correct, the probability of getting this result (or a more extreme result) is very low -- thus we reject H0 and accept H1 Hypothesis testing IV Probailities The meaning of p What does p means? What is the difference between p = 0.03, p = 0.001, & p = 0.11 What is the relationship between p and conﬁdence? What is the relationship between p and effect size? What is the relationship between p and number of subjects? 2 types of errors Not conservative and liberal just balancing 2 types of error H0 is wrong Correct Type 2 error H0 is correct Type I error Reject H0 Correct Accept H0 The importance of effect size Always give effect size measures Mean difference Quartile differences etc. Summary Hypotheses testing H1 & H0 The meaning of p 2 types of errors Effect size ! Within & between subject designs