Selective trials

A principal-agent approach to randomized controlled experiments

Sylvain Chassang, Gerard Padró I Miquel, Erik Snowberg

    Research output: Contribution to journalReview article

    Abstract

    We study the design of randomized controlled experiments when outcomes are significantly affected by experimental subjects' unobserved effort expenditure. While standard randomized controlled trials (RCTs) are internally consistent, the unobservability of effort compromises external validity. We approach trial design as a principal-agent problem and show that natural extensions of RCTs-which we call selective trials-can help improve external validity. In particular, selective trials can disentangle the effects of treatment, effort, and the interaction of treatment and effort. Moreover, they can help identify when treatment effects are affected by erroneous beliefs and inappropriate effort expenditure.

    Original languageEnglish (US)
    Pages (from-to)1279-1309
    Number of pages31
    JournalAmerican Economic Review
    Volume102
    Issue number4
    DOIs
    StatePublished - Jun 2012

    Fingerprint

    Experiment
    Randomized controlled trial
    External validity
    Expenditure
    Compromise
    Interaction
    Treatment effects
    Principal-agent problem

    ASJC Scopus subject areas

    • Economics and Econometrics

    Cite this

    Selective trials : A principal-agent approach to randomized controlled experiments. / Chassang, Sylvain; Padró I Miquel, Gerard; Snowberg, Erik.

    In: American Economic Review, Vol. 102, No. 4, 06.2012, p. 1279-1309.

    Research output: Contribution to journalReview article

    Chassang, Sylvain ; Padró I Miquel, Gerard ; Snowberg, Erik. / Selective trials : A principal-agent approach to randomized controlled experiments. In: American Economic Review. 2012 ; Vol. 102, No. 4. pp. 1279-1309.
    @article{48f008e02fcc468c94981d243d4d72e5,
    title = "Selective trials: A principal-agent approach to randomized controlled experiments",
    abstract = "We study the design of randomized controlled experiments when outcomes are significantly affected by experimental subjects' unobserved effort expenditure. While standard randomized controlled trials (RCTs) are internally consistent, the unobservability of effort compromises external validity. We approach trial design as a principal-agent problem and show that natural extensions of RCTs-which we call selective trials-can help improve external validity. In particular, selective trials can disentangle the effects of treatment, effort, and the interaction of treatment and effort. Moreover, they can help identify when treatment effects are affected by erroneous beliefs and inappropriate effort expenditure.",
    author = "Sylvain Chassang and {Padr{\'o} I Miquel}, Gerard and Erik Snowberg",
    year = "2012",
    month = "6",
    doi = "10.1257/aer.102.4.1279",
    language = "English (US)",
    volume = "102",
    pages = "1279--1309",
    journal = "American Economic Review",
    issn = "0002-8282",
    publisher = "American Economic Association",
    number = "4",

    }

    TY - JOUR

    T1 - Selective trials

    T2 - A principal-agent approach to randomized controlled experiments

    AU - Chassang, Sylvain

    AU - Padró I Miquel, Gerard

    AU - Snowberg, Erik

    PY - 2012/6

    Y1 - 2012/6

    N2 - We study the design of randomized controlled experiments when outcomes are significantly affected by experimental subjects' unobserved effort expenditure. While standard randomized controlled trials (RCTs) are internally consistent, the unobservability of effort compromises external validity. We approach trial design as a principal-agent problem and show that natural extensions of RCTs-which we call selective trials-can help improve external validity. In particular, selective trials can disentangle the effects of treatment, effort, and the interaction of treatment and effort. Moreover, they can help identify when treatment effects are affected by erroneous beliefs and inappropriate effort expenditure.

    AB - We study the design of randomized controlled experiments when outcomes are significantly affected by experimental subjects' unobserved effort expenditure. While standard randomized controlled trials (RCTs) are internally consistent, the unobservability of effort compromises external validity. We approach trial design as a principal-agent problem and show that natural extensions of RCTs-which we call selective trials-can help improve external validity. In particular, selective trials can disentangle the effects of treatment, effort, and the interaction of treatment and effort. Moreover, they can help identify when treatment effects are affected by erroneous beliefs and inappropriate effort expenditure.

    UR - http://www.scopus.com/inward/record.url?scp=84862603272&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84862603272&partnerID=8YFLogxK

    U2 - 10.1257/aer.102.4.1279

    DO - 10.1257/aer.102.4.1279

    M3 - Review article

    VL - 102

    SP - 1279

    EP - 1309

    JO - American Economic Review

    JF - American Economic Review

    SN - 0002-8282

    IS - 4

    ER -