Replication & Reproducibility

Issues, Concepts and an Introduction to Code Versioning

What is the replicability crisis? How widespread is it? Where does it come from, and what can we do about it? This seminar covers the fundamentals of the replication crisis and the current debate around the tools used to navigate it, from statistical roots and questionable research practices to preregistration, registered reports, and code versioning.

At a glance

  • Programme: Psychologischer Wahlbereich (B), MSc Psychology, Institute of Psychology, Universität Hamburg
  • Instructor: Prof. Schuck
  • Format: weekly 90-minute seminar during the winter term

Requirements

  • Read the assigned research articles each week
  • Give one 30-min presentation in class, followed by a 30-min discussion
  • Submit one question to the presenter before each class

Presentation guidelines

Each presentation (30–35 min, two presenters per session) should cover:

  • Relevant background and motivation — what the question is and why it is interesting
  • The study design and methodological aspects — for empirical studies: number of subjects, task description, conditions, duration; for simulations: simulation details — answering what was done
  • An explanation of any concepts that may be unfamiliar to your colleagues
  • Hypotheses — what specifically is expected in the data
  • Results — what they are and how they were obtained
  • A one-slide summary that flags open questions or issues

Discussion guidelines

Each discussion (25–30 min) should be an interactive in-class session that engages with the topic of the presentation and uses the submitted question as input. For example: round tables for different questions, hands-on demonstrations, short presentations, or discussion of a related paper.

Sessions & readings

Session 1

Background and foundations

Understand the basic premises of modern science.

  • Chalmers, A. F. (2013). Falsificationism and progress. In What is this thing called science? (4th ed., pp. 66–83). Hackett Publishing.
  • National Academies of Sciences, Engineering, and Medicine. (2019). Reproducibility and replicability in science (Executive Summary). The National Academies Press.
Session 2

Statistical roots

Understand what effect power, bias, and flexible analyses have on the rate of false positives and false negatives.

Session 3

Empirical evidence of a "crisis" in psychology (and elsewhere)

Get a sense for the scale and scope of the reproducibility crisis.

Session 4

Case study: Does putting a pen in your mouth make you happy?

Understand how the discussion is shaped by multiple perspectives.

Session 5

Computational reproducibility

Understand the issues that complicate reproducing computational pipelines.

Session 6

Impact of post-hoc decisions

Identify p-hacking, HARKing, and selective reporting; understand systemic incentives.

Session 7

Research culture and beliefs around reproducibility

Which incentive structures could play a role in science?

Session 8

Preregistration, registered reports & replications

Different ways to do science and publish results — preregistration, preprints, eLife, executable notebooks.

Session 9

Changing p-values

Understand the effect of changing the p-value standard on the literature.

  • Benjamin, D. J., Berger, J. O., Johannesson, M., Nosek, B. A., Wagenmakers, E. J., et al. (2018). Redefine statistical significance. Nature Human Behaviour, 2(1), 6–10.
Session 10

Bayesian approaches; AIC vs. BIC

Evaluate Bayesian approaches as alternatives.

Session 11

Sample sizes

Session 12

Hands-on example

Analyze one example data set and look at consistency and agreement — a p-hacking contest.

Session 13

Hands-on: Git, OSF, containers & Open Science

Practical introduction to code versioning and open-science tooling.

Session 14

Discussion

Short take-away presentations: your most important lesson from the course, or something that wasn't covered.

Back to Home