Conference announcement

The activity of statistics as conventionally conceived is a rather solitary affair. It’s just you and the data, and you have to make some decision based on the data - estimate a parameter, test a hypothesis, etc. In practice, however, statistics (and empirical research more generally) is a social endeavor, involving different researchers, journal editors and referees, readers, policymakers, and others. This social dimension of statistics has come to the fore in recent debates about publication bias, p-hacking, and the perceived replicability crisis of various empirical fields.

Taking the social dimension seriously suggests an alternative perspective on statistics, where the task of empirical researchers is to provide useful summaries of complex data to their readers in order to promote some form of collective learning. This task is subject to costs of time and attention of researchers, referees, and readers, as well as constraints on social learning in terms of limited information, strategic behavior, the sociology of research, etc.

The goal of this conference is to bring together researchers from various fields, including econometrics and statistics, economic theory, empirical practice, and the history, sociology, and philosophy of science, in order to promote a discussion of statistical methods in their social context.

Possible topics for this conference include, but are not restricted to:

  • Publication bias and p-hacking
  • Replicability and replications
  • Pre-analysis plans and other reform proposals
  • Communication, limited attention, value of information, social learning
  • Paradigm changes, the role of method
  • The sociology of empirical research

This conference is supported by the Alfred P. Sloan Foundation grant “Publication bias and specification searching. Identification, correction, and reform proposals.”

Schedule

Friday, May 10

8:30am Breakfast

9:00am Maximilian Kasy (econometrics)
Opening remarks Slides

9:30am Katherine Casey (development economics)
Comments on pre-specification and analysis plans manuscript 1, manuscript 2

10:15am Simine Vazire (psychology)
The Credibility Revolution in Psychological Science

11:00am Break

11:15am Ben Olken (development economics)
Promises and Perils of Pre-Analysis Plans manuscript

12:00pm Jann Spiess (econometrics)
Optimal Estimation when Researcher and Social Preferences are Misaligned manuscript

12:45pm Lunch

1:45pm Alex Frankel (economic theory)
Which findings should be published manuscript

2:30pm Isaiah Andrews (econometrics)
Statistical Reports for Remote Agents manuscript

3:15pm Break

3:45pm Marco Ottaviani (economic theory)
Strategic Sample Selection manuscript

4:30pm Break

7pm Dinner at Cambridge One

Saturday, May 11

8:30am Breakfast

9:00am Theodore Porter (history of statistics)
Statistics, a Tool of Science? abstract

9:45am Deborah Mayo (philosophy of statistics)
3D Statistics: 7 Responses to Challenges for Statistical Testers chapter

10:45am Break

11am Zoe Hitzig (economic theory, philosophy)
The Problem of New Evidence: P-hacking and Pre-analysis Plans

11:45am Daniel Mellow (meta studies)

12:15pm Adjourn

Logistics

  • Date: May 10-11, 2019
  • Location: Department of Economics, Harvard University
    3rd floor lounge, Littauer Hall, 1805 Cambridge Street, Cambridge, MA 02138
  • Presentations: Target length 30 minutes, with 15 minutes discussion
  • Reimbursement: We will provide a honorarium of 800$ for out-of-town participants, to cover hotel and travel costs.
  • Contact: For logistical questions, if you would like us to reserve a hotel for you, and for honararium payments, please contact Trina Ott.
  • Papers and Slides: Please also email these to Trina, and we will post them on this website before the conference.
  • Dinner for speakers on May 10 will take place at Cambridge One, 27 Church Street, Cambridge.
  • Videos of talks will be recorded and posted online; please let us know if you don’t want to be recorded.

Confirmed participants

  • Isaiah Andrews (econometrics) Statistical Reports for Remote Agents manuscript
  • Katherine Casey (development economics) Comments on pre-specification and analysis plans manuscript 1, manuscript 2
  • Alex Frankel (economic theory) Which findings should be published manuscript
  • Zoe Hitzig (economic theory, philosophy) The Problem of New Evidence: P-hacking and Pre-analysis Plans
  • Maximilian Kasy (econometrics) Opening remarks Slides
  • Deborah Mayo (philosophy of statistics) 3D Statistics: 7 Responses to Challenges for Statistical Testers chapter
  • Ben Olken (development economics) Promises and Perils of Pre-Analysis Plans manuscript
  • Marco Ottaviani (economic theory) Strategic Sample Selection manuscript
  • Theodore Porter (history of statistics) Statistics, a Tool of Science? abstract
  • Jann Spiess (econometrics) Optimal Estimation when Researcher and Social Preferences are Misaligned manuscript
  • Simine Vazire (psychology) The Credibility Revolution in Psychological Science