Dorothy V. M. Bishop University of

Slides:



Advertisements
Similar presentations
SAMPLING. Next week  2 book chapters  Outline of thesis proposal/paper intro  Find a scale and answer questions  Thought paper.
Advertisements

Pre-analysis plans Module 8.3. Recap on statistics If we find a result is significant at the 5% level, what does this mean? – there is a 5% or less probability.
Statistical Significance What is Statistical Significance? What is Statistical Significance? How Do We Know Whether a Result is Statistically Significant?
HYPOTHESIS TESTING Four Steps Statistical Significance Outcomes Sampling Distributions.
Statistical Significance What is Statistical Significance? How Do We Know Whether a Result is Statistically Significant? How Do We Know Whether a Result.
Click on image for full.pdf article Links in article to access datasets.
Today Concepts underlying inferential statistics
Osama A Samarkandi, PhD-RN, NIAC BSc, GMD, BSN, MSN.
© 2012 Cengage Learning. All Rights Reserved. May not be scanned, copied or duplicated, or posted to a publicly accessible website, in whole or in part.
Educating Students in the Practice of Psychological Science Robert W. Proctor E. J. Capaldi Gregory Francis May 27, 2012 Symposium: Reforming Psychology's.
School Counselors Doing Action Research Jay Carey and Carey Dimmitt Center for School Counseling Outcome Research UMass Amherst CT Guidance Leaders March.
One-Way Manova For an expository presentation of multivariate analysis of variance (MANOVA). See the following paper, which addresses several questions:
Academic Viva POWER and ERROR T R Wilson. Impact Factor Measure reflecting the average number of citations to recent articles published in that journal.
Main issues Effect-size ratio Development of protocols and improvement of designs Research workforce and stakeholders Reproducibility practices and reward.
Educational Research: Competencies for Analysis and Application, 9 th edition. Gay, Mills, & Airasian © 2009 Pearson Education, Inc. All rights reserved.
The Nature of Science Chapter 1: What is Science?
Introduction to Statistics Osama A Samarkandi, PhD, RN BSc, GMD, BSN, MSN, NIAC Deanship of Skill development Dec. 2 nd -3 rd, 2013.
IMPORTANCE OF STATISTICS MR.CHITHRAVEL.V ASST.PROFESSOR ACN.
Replication in Prevention Science Valentine, et al.
Doing the Right Thing! … statistically speaking...
BIOL 582 Lecture Set 2 Inferential Statistics, Hypotheses, and Resampling.
April Center for Open Fostering openness, integrity, and reproducibility of scientific research.
Chapter 22 Inferential Data Analysis: Part 2 PowerPoint presentation developed by: Jennifer L. Bellamy & Sarah E. Bledsoe.
Copyright © Allyn & Bacon 2007 Chapter 2 Research Methods This multimedia product and its contents are protected under copyright law. The following are.
Looking for statistical twins
Scottish National Burden of Disease, Injuries and Risk Factors study:
Unit 1 Lesson 3 Scientific Investigations
Statistics & Evidence-Based Practice
Advanced Data Analytics
AP Seminar: Statistics Primer
Chapter 2: The Research Enterprise in Psychology
DATA COLLECTION METHODS IN NURSING RESEARCH
Selecting the Best Measure for Your Study
Statistics in Clinical Trials: Key Concepts
Reproducibility Project: Psychology A Discussion
How to Critically Appraise Literature
Reporting quality in preclinical studies Emily S Sena, PhD Centre for Clinical Brain Sciences, University of
Introduction to Statistics for Engineers
Introduction to Statistics: Probability and Types of Analysis
AP Seminar: Statistics Primer
Observational Study vs. Experimental Design
Hypothesis Testing and Confidence Intervals (Part 1): Using the Standard Normal Lecture 8 Justin Kern October 10 and 12, 2017.
Statistical Data Analysis
Meta-Analytic Thinking
Study Pre-Registration
Regression Statistics
A2 unit 4 Clinical Psychology
Calculating Sample Size: Cohen’s Tables and G. Power
Building a GER Toolbox As you return from break, please reassemble into your working groups: Surveys and Instruments Analytical Tools Getting Published.
Two-sided p-values (1.4) and Theory-based approaches (1.5)
Session 2 Challenges and benefits of teaching controversial issues
R.J.Watt D.I.Donaldson University of Stirling
Critical Appraisal & Literature review
UNDERSTANDING RESEARCH RESULTS: STATISTICAL INFERENCE
Psych 231: Research Methods in Psychology
Introduction: Statistics meets corpus linguistics
Statistical Data Analysis
Chapter 8 Making Sense of Statistical Significance: Effect Size, Decision Errors, and Statistical Power.
Psych 231: Research Methods in Psychology
Psych 231: Research Methods in Psychology
What are systematic reviews and why do we need them?
Copyright © Allyn & Bacon 2007
Chapter 9: Significance Testing
Chapter 15 Analysis of Variance
Chapter 4 Summary.
Unit 1: Scientific Inquiry
Meta-analysis, systematic reviews and research syntheses
Maternal Factors of Childhood Obesity
Critical Appraisal & Literature review
Open Science & Reproducibility
Presentation transcript:

Dorothy V. M. Bishop University of Oxford @deevybee Hacking a way through the garden of forking paths: A cause of poor reproducibility Dorothy V. M. Bishop University of Oxford @deevybee

This not new Failure to distinguish between hypothesis-testing and hypothesis-generating (exploratory) research -> misuse of statistical tests ‘If the processing of empirically obtained material has in any way an “exploratory character”, i.e. if the attempt to let the material speak leads to ad hoc decisions in terms of processing, as described above, then this precludes the exact interpretability of possible outcomes of statistical tests’. De Groot, 1956 de Groot, A. D. (2014). The meaning of “significance” for different types of research [translated and annotated by EJ Wagenmakers et al]. Acta Psychologica, 148, 188-194

Writing the Empirical Journal Article Daryl J. Bem The Compleat Academic: A Practical Guide for the Beginning Social Scientist, 2nd Edition. Washington, DC: American Psychological Association, 2004. Which Article Should You Write? There are two possible articles you can write: (a) the article you planned to write when you designed your study or (b) the article that makes the most sense now that you have seen the results. They are rarely the same, and the correct answer is (b). re Data Analysis: Examine them from every angle. Analyze the sexes separately. Make up new composite indexes. If a datum suggests a new hypothesis, try to find additional evidence for it elsewhere in the data. If you see dim traces of interesting patterns, try to reorganize the data to bring them into bolder relief. If there are participants you don’t like, or trials, observers, or interviewers who gave you anomalous results, drop them (temporarily). Go on a fishing expedition for something— anything —interesting. “This book provides invaluable guidance that will help new academics plan, play, and ultimately win the academic career game.”

Large population database used to explore link between ADHD and handedness 1 contrast Probability of a ‘significant’ p-value < .05 = .05

Large population database used to explore link between ADHD and handedness Focus just on Young subgroup: 2 contrasts at this level Probability of a ‘significant’ p-value < .05 = .10

Large population database used to explore link between ADHD and handedness Focus just on Young on measure of hand skill: 4 contrasts at this level Probability of a ‘significant’ p-value < .05 = .19

Large population database used to explore link between ADHD and handedness Focus just on Young, Females on measure of hand skill: 8 contrasts at this level Probability of a ‘significant’ p-value < .05 = .34

Large population database used to explore link between ADHD and handedness Focus just on Young, Urban, Females on measure of hand skill: 16 contrasts at this level Probability of a ‘significant’ p-value < .05 = .56

Problem exacerbated because Can now easily gather huge multivariate datasets Can easily do complex statistical analyses Problems with exploratory analyses that use methods that presuppose hypothesis-testing approach

Huge bias for type I error

Solutions a. Using simulated datasets to give insight into statistical methods

Correlation matrix: 8 random normal deviates

Correlation matrix: 8 random normal deviates

Correlation matrix: 8 random normal deviates

Correlation matrix: 8 random normal deviates

We must overcome our natural bias to over-interpret observed patterns

Another example: Multiway ANOVA Illustrated with field of ERP/EEG Flexibility in analysis in terms of: Electrodes Time intervals Frequency ranges Measurement of peaks etc, etc Often see analyses with 4- or 5-way ANOVA (group x side x site x condition x interval) Standard stats packages correct p-values for N levels WITHIN a factor, but not for overall N factors and interactions . Cramer AOJ, et al 2016. Hidden multiplicity in exploratory multiway ANOVA: Prevalence and remedies. Psychonomic Bulletin & Review 23:640-647

Solutions b. Distinguish exploration from hypothesis-testing analyses Subdivide data into exploration and replication sets. Or replicate in another dataset

Solutions c. Preregistration of analyses

Solutions for journals More emphasis on methodological rigour Publish well-powered null findings/replications http://deevybee.blogspot.co.uk/

Solutions for institutions Problems caused by employers Reward research reproducibility over impact factor in evaluation Consider ‘bang for your buck’ rather than amount of grant income Reward those who adopt open science practices Solutions for institutions Marcia McNutt Science 2014 • VOL 346 ISSUE 6214 "This is Dr Bagshaw, discoverer of the infinitely expanding research grant“ ©Cartoonstock

Solutions for funders Problems caused by funders Promote data-sharing; require that all data are reported Fund well-designed replications of key findings Encourage pre-registration More scrutiny of methods – can’t assume competence

Resources http://www.acmedsci.ac.uk/policy/policy-projects/reproducibility-and-reliability-of-biomedical-research/symposium-resources-links/ http://www.slideshare.net/deevybishop/references-on-reproducibility-crisis-in-science-by-dvm-bishop