Section outline

  • P-values and confidence intervals,  Alternatives to randomized trials

    Introduction and justification: why this Bayesian stuff is important and why we think you can handle it; what P-values and confidence intervals don' t mean; what they do mean; confidence intervals for negative studies and for proportions with small numerators.  Alternatives to RCTs: When randomized trials are and may not be needed; instrumental variables and natural experiments; measuring additonal predictor and outcome variables to estimate bias; propensity scores.

    Faculty:  Thomas Newman

    • Lecture Slides:

    • Lec 9 Alternatives To RCTs and P-values 2021-1115 File

    • Lecture Audio/Video Recording (Access restricted to registered students):

    • Tom's 2019 Lecture on Alternatives to RCTS and P-values and CI (WATCH FOR 2020) URL
      Update for Slide 28: Vinay Prasad is in our department at UCSF now!
      Not available unless: You belong to Registered students only

    • Large Group Discussion (Access restricted to registered students):
      Brief formal review of lecture followed by question and answer discussion. Recorded lecture should be viewed prior to this session.

    • Zoom Recording of Large Group 11/16/2021 Media Resource
      Not available unless: You belong to Registered students only
    • Chapter 9 and 11 ILG 2021-1115 File

    • Required Reading:

      EBD-2, Ch 9 and 11

    • Goodman S 2008 Seminars Hematol Dirty Dozen P Value fallacies File
      Not available unless: Your ID number contains 02

    • Optional Reading:

    • Laptook 2017 JAMA Effect of therapeutic hypothermia after 6 hours Bayesian example File
      Not available unless: Your ID number contains 02
    • Quintana 2017 JAMA Bayesian analysis. Using prior information to interpret results of clinical trials File
      Not available unless: Your ID number contains 02
    • Browner WS, Newman TB. Are all significant P-values created equal? The analogy between diagnostic tests and clinical research. JAMA 1987;257:2459-63 File
      Not available unless: You belong to a group in Registered Students Only
    • Goodman, S. Toward evidence-based medical statistics. 1: The P value fallacy. Annals of Internal Medicine 1999; 130: 995-1004 File
      Not available unless: You belong to a group in Registered Students Only
    • Goodman, S. Toward evidence-based medical statistics. 2: The Bayes factor. Annals of Internal Medicine 1999; 130: 1005-1013 File
      Not available unless: You belong to a group in Registered Students Only
    • Newman, T. If almost nothing goes wrong, is almost everything all right? JAMA 1995; 274: 1013 File
      Not available unless: You belong to a group in Registered Students Only
    • Goodman S 2013 Circulation Are we there yet File
      Not available unless: Your ID number contains 02
    • Ray 2006 American Journal of Epidemiology New User DEsigns File
      Not available unless: Your ID number contains 02
    • Redelmeier 2001 AIM Survival in academy arard winners File
    • Sylvestre 2006 AIM Do Oscar winners live longer File

    • Assignment:  Problem set #9 - Alternatives to RCTs; P- values and confidence intervals


    • Problem Set #9 File

    • Assignment Due Date: 11/19/2020 at 1:15 PM

    • Problem Set #9 Upload Assignment


    • Problem Set #9 (M.Laker Section) Assignment
      Not available unless: You belong to M. Laker

    • Assignment Answer Key (Access restricted to registered students):

    • PSet9 (chs 9 and 11) with Answers and changes 2021.1103.docx File
    • Newman 2021 JAMA IM LTE Possible Immortal Time Bias in Study of Antibiotics and Asthma File
    • Newman 2021 JAMA IM Antibiotics and asthma what have we learned Immortal Time Bias File