Printable Handouts

PDF

Navigable Slide Index

1. Introduction
2. Three questions
3. Hypothesis testing (1)
4. Hypothesis testing (2)
5. Significance testing (1)
6. Significance testing (2)
7. Bayesian inference (1)
8. Bayesian inference (2)
9. A diagnostic test (1)
10. A diagnostic test (2)
11. The law of likelihood
12. The law says
13. Degrees of strength
14. Three categories of the strength of evidence
15. Diagnostic test, revisited
16. A positive test result and disease presence
17. Irrelevant for data interpretation
18. The diagnostic test
19. Another diagnostic test
20. Question
21. Answer
22. An observed positive result can be misleading
23. The key to the argument
24. Three key concepts (1)
25. Three key concepts (2)
26. Old news
27. The University Group Diabetes Program
28. UGDP data
29. Binomial likelihood
30. Probability of cardiovascular death: placebo (1)
31. Probability of cardiovascular death: placebo (2)
32. Probability of cardiovascular death: tolbutamide (1)
33. Probability of cardiovascular death: placebo (3)
34. Probability of cardiovascular death: tolbutamide (2)
35. Relative risk of cardiovascular death (1)
36. Relative risk of cardiovascular death (2)
37. 'Undesirable' evidence
38. Design considerations
39. Probabilities of misleading and weak evidence (1)
40. Probabilities of misleading and weak evidence (2)
41. Efficiency of sequential designs
42. Simulation
43. Results
44. The universal bound
45. Universal bound holds for sequential designs
46. The bump function
47. Bump equations
48. Probabilities of generating misleading evidence (1)
49. The tepee function
50. Tepee equations
51. Probabilities of generating misleading evidence (2)
52. Probabilities of generating misleading evidence (3)
53. Composite hypothesis (1)
54. Composite hypothesis (2)
55. Relative risk of cardiovascular death (3)
56. So what?
57. Extensions
58. References

Topics Covered

• Using likelihood ratios to measure the strength of statistical evidence in data
• Paradigms for measuring statistical evidence: the three questions
• Likelihood ratios
• The law of likelihood
• The likelihood principle
• Misleading evidence
• Probabilities of observing misleading evidence and their bounds
• Sequential designs in the likelihood paradigm
• Composite hypothesis and extensions
• Example using real data from a well known clinical trial

Links

Series:

• Design and Analysis of Randomized Clinical Trials

Categories:

• Methods
• Pharmaceutical Sciences

Talk Citation

Blume, J. (2017, September 28). Likelihood ratios and the strength of statistical evidence [Video file]. In The Biomedical & Life Sciences Collection, Henry Stewart Talks. Retrieved November 21, 2024, from https://doi.org/10.69645/JOSG3257.
Export Citation (RIS)

Publication History

• Published on October 1, 2007
• Reviewed on September 28, 2017

Financial Disclosures

• Prof. Jeffrey Blume has not informed HSTalks of any commercial/financial relationship that it is appropriate to disclose.