Statistical treatment of looking-time data

G. Csibra, Mikolaj Hernik, Olivier Mascaro, Denis Tatone, Máté Lengyel

Research output: Contribution to journalArticle

42 Citations (Scopus)

Abstract

Looking times (LTs) are frequently measured in empirical research on infant cognition. We analyzed the statistical distribution of LTs across participants to develop recommendations for their treatment in infancy research. Our analyses focused on a common within-subject experimental design, in which longer looking to novel or unexpected stimuli is predicted. We analyzed data from 2 sources: an in-house set of LTs that included data from individual participants (47 experiments, 1,584 observations), and a representative set of published articles reporting group-level LT statistics (149 experiments from 33 articles). We established that LTs are log-normally distributed across participants, and therefore, should always be log-transformed before parametric statistical analyses. We estimated the typical size of significant effects in LT studies, which allowed us to make recommendations about setting sample sizes. We show how our estimate of the distribution of effect sizes of LT studies can be used to design experiments to be analyzed by Bayesian statistics, where the experimenter is required to determine in advance the predicted effect size rather than the sample size. We demonstrate the robustness of this method in both sets of LT experiments.

Original languageEnglish
Pages (from-to)521-536
Number of pages16
JournalDevelopmental Psychology
Volume52
Issue number4
DOIs
Publication statusPublished - Apr 1 2016

Fingerprint

Time and Motion Studies
Sample Size
experiment
Therapeutics
Statistical Distributions
Empirical Research
Bayesian statistics
Information Storage and Retrieval
Cognition
time
Research Design
empirical research
cognition
infant
stimulus
statistics
Research
Group

Keywords

  • Bayesian statistics
  • Infancy
  • Log-normal distribution
  • Looking times

ASJC Scopus subject areas

  • Developmental and Educational Psychology
  • Life-span and Life-course Studies
  • Demography

Cite this

Csibra, G., Hernik, M., Mascaro, O., Tatone, D., & Lengyel, M. (2016). Statistical treatment of looking-time data. Developmental Psychology, 52(4), 521-536. https://doi.org/10.1037/dev0000083

Statistical treatment of looking-time data. / Csibra, G.; Hernik, Mikolaj; Mascaro, Olivier; Tatone, Denis; Lengyel, Máté.

In: Developmental Psychology, Vol. 52, No. 4, 01.04.2016, p. 521-536.

Research output: Contribution to journalArticle

Csibra, G, Hernik, M, Mascaro, O, Tatone, D & Lengyel, M 2016, 'Statistical treatment of looking-time data', Developmental Psychology, vol. 52, no. 4, pp. 521-536. https://doi.org/10.1037/dev0000083
Csibra G, Hernik M, Mascaro O, Tatone D, Lengyel M. Statistical treatment of looking-time data. Developmental Psychology. 2016 Apr 1;52(4):521-536. https://doi.org/10.1037/dev0000083
Csibra, G. ; Hernik, Mikolaj ; Mascaro, Olivier ; Tatone, Denis ; Lengyel, Máté. / Statistical treatment of looking-time data. In: Developmental Psychology. 2016 ; Vol. 52, No. 4. pp. 521-536.
@article{1fe996fc729b4a1ba6d87590746d36c0,
title = "Statistical treatment of looking-time data",
abstract = "Looking times (LTs) are frequently measured in empirical research on infant cognition. We analyzed the statistical distribution of LTs across participants to develop recommendations for their treatment in infancy research. Our analyses focused on a common within-subject experimental design, in which longer looking to novel or unexpected stimuli is predicted. We analyzed data from 2 sources: an in-house set of LTs that included data from individual participants (47 experiments, 1,584 observations), and a representative set of published articles reporting group-level LT statistics (149 experiments from 33 articles). We established that LTs are log-normally distributed across participants, and therefore, should always be log-transformed before parametric statistical analyses. We estimated the typical size of significant effects in LT studies, which allowed us to make recommendations about setting sample sizes. We show how our estimate of the distribution of effect sizes of LT studies can be used to design experiments to be analyzed by Bayesian statistics, where the experimenter is required to determine in advance the predicted effect size rather than the sample size. We demonstrate the robustness of this method in both sets of LT experiments.",
keywords = "Bayesian statistics, Infancy, Log-normal distribution, Looking times",
author = "G. Csibra and Mikolaj Hernik and Olivier Mascaro and Denis Tatone and M{\'a}t{\'e} Lengyel",
year = "2016",
month = "4",
day = "1",
doi = "10.1037/dev0000083",
language = "English",
volume = "52",
pages = "521--536",
journal = "Developmental Psychology",
issn = "0012-1649",
publisher = "American Psychological Association Inc.",
number = "4",

}

TY - JOUR

T1 - Statistical treatment of looking-time data

AU - Csibra, G.

AU - Hernik, Mikolaj

AU - Mascaro, Olivier

AU - Tatone, Denis

AU - Lengyel, Máté

PY - 2016/4/1

Y1 - 2016/4/1

N2 - Looking times (LTs) are frequently measured in empirical research on infant cognition. We analyzed the statistical distribution of LTs across participants to develop recommendations for their treatment in infancy research. Our analyses focused on a common within-subject experimental design, in which longer looking to novel or unexpected stimuli is predicted. We analyzed data from 2 sources: an in-house set of LTs that included data from individual participants (47 experiments, 1,584 observations), and a representative set of published articles reporting group-level LT statistics (149 experiments from 33 articles). We established that LTs are log-normally distributed across participants, and therefore, should always be log-transformed before parametric statistical analyses. We estimated the typical size of significant effects in LT studies, which allowed us to make recommendations about setting sample sizes. We show how our estimate of the distribution of effect sizes of LT studies can be used to design experiments to be analyzed by Bayesian statistics, where the experimenter is required to determine in advance the predicted effect size rather than the sample size. We demonstrate the robustness of this method in both sets of LT experiments.

AB - Looking times (LTs) are frequently measured in empirical research on infant cognition. We analyzed the statistical distribution of LTs across participants to develop recommendations for their treatment in infancy research. Our analyses focused on a common within-subject experimental design, in which longer looking to novel or unexpected stimuli is predicted. We analyzed data from 2 sources: an in-house set of LTs that included data from individual participants (47 experiments, 1,584 observations), and a representative set of published articles reporting group-level LT statistics (149 experiments from 33 articles). We established that LTs are log-normally distributed across participants, and therefore, should always be log-transformed before parametric statistical analyses. We estimated the typical size of significant effects in LT studies, which allowed us to make recommendations about setting sample sizes. We show how our estimate of the distribution of effect sizes of LT studies can be used to design experiments to be analyzed by Bayesian statistics, where the experimenter is required to determine in advance the predicted effect size rather than the sample size. We demonstrate the robustness of this method in both sets of LT experiments.

KW - Bayesian statistics

KW - Infancy

KW - Log-normal distribution

KW - Looking times

UR - http://www.scopus.com/inward/record.url?scp=84979650484&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979650484&partnerID=8YFLogxK

U2 - 10.1037/dev0000083

DO - 10.1037/dev0000083

M3 - Article

C2 - 26845505

AN - SCOPUS:84979650484

VL - 52

SP - 521

EP - 536

JO - Developmental Psychology

JF - Developmental Psychology

SN - 0012-1649

IS - 4

ER -