A step forward: Bayesian hierarchical modelling as a tool in assessment of individual discrimination performance

Maartje de Klerk; Duco Veen; Frank Wijnen; Elise de Bree

doi:10.1016/j.infbeh.2019.101345

A step forward: Bayesian hierarchical modelling as a tool in assessment of individual discrimination performance

Maartje de Klerk^*, Duco Veen, Frank Wijnen, Elise de Bree

^*Corresponding author for this work

University of Amsterdam

Research output: Contribution to journal › Article › Academic › peer-review

Abstract

Individual assessment of infants’ speech discrimination is of great value for studies of language development that seek to relate early and later skills, as well as for clinical work. The present study explored the applicability of the hybrid visual fixation paradigm (Houston et al., 2007) and the associated statistical analysis approach to assess individual discrimination of a native vowel contrast, /aː/ - /eː/, in Dutch 6 to 10-month-old infants. Houston et al. found that 80% (8/10) of the 9-month-old infants successfully discriminated the contrast between pseudowords boodup - seepug. Using the same approach, we found that 12% (14/117) of the infants in our sample discriminated the highly salient /aː/ -/eː/ contrast. This percentage was reduced to 3% (3/117) when we corrected for multiple testing. Bayesian hierarchical modeling indicated that 50% of the infants showed evidence of discrimination. Advantages of Bayesian hierarchical modeling are that 1) there is no need for a correction for multiple testing and 2) better estimates at the individual level are obtained. Thus, individual speech discrimination can be more accurately assessed using state of the art statistical approaches.

Original language	English
Article number	101345
Journal	Infant Behavior and Development
Volume	57
DOIs	https://doi.org/10.1016/j.infbeh.2019.101345
Publication status	Published - 1 Nov 2019

Funding

We are grateful to the infants and their caregivers for participating. We would like to thank the student assistants Sule Kurtçebe, Tinka Versteegh, Lorijn Zaadnoordijk and Joleen Zuidema, who helped collecting data. We would like to thank Annemarie Kerkhoff for her help in the design of the experiment and Derek Houston for sharing some of his raw data with us (see Appendix A ). This research was funded by The Netherlands Organization for Scientific Research (NWO) . Grants nr. 360-70-270 , awarded to F.N.K. Wijnen and nr. VIDI-452-14-006 , awarded to R. van de Schoot. Appendix A See . Table A1

Keywords

Autoregressive error structure
Bayesian hierarchical modeling
Hybrid visual fixation
Individual analysis
Speech sound discrimination

Access to Document

10.1016/j.infbeh.2019.101345

1-s2.0-S0163638318300869-mainFinal published version, 6.01 MBLicence: Taverne

Cite this

@article{04c8264d09444699908c8547f5fb4383,

title = "A step forward: Bayesian hierarchical modelling as a tool in assessment of individual discrimination performance",

abstract = "Individual assessment of infants{\textquoteright} speech discrimination is of great value for studies of language development that seek to relate early and later skills, as well as for clinical work. The present study explored the applicability of the hybrid visual fixation paradigm (Houston et al., 2007) and the associated statistical analysis approach to assess individual discrimination of a native vowel contrast, /aː/ - /eː/, in Dutch 6 to 10-month-old infants. Houston et al. found that 80% (8/10) of the 9-month-old infants successfully discriminated the contrast between pseudowords boodup - seepug. Using the same approach, we found that 12% (14/117) of the infants in our sample discriminated the highly salient /aː/ -/eː/ contrast. This percentage was reduced to 3% (3/117) when we corrected for multiple testing. Bayesian hierarchical modeling indicated that 50% of the infants showed evidence of discrimination. Advantages of Bayesian hierarchical modeling are that 1) there is no need for a correction for multiple testing and 2) better estimates at the individual level are obtained. Thus, individual speech discrimination can be more accurately assessed using state of the art statistical approaches.",

keywords = "Autoregressive error structure, Bayesian hierarchical modeling, Hybrid visual fixation, Individual analysis, Speech sound discrimination",

author = "{de Klerk}, Maartje and Duco Veen and Frank Wijnen and {de Bree}, Elise",

year = "2019",

month = nov,

day = "1",

doi = "10.1016/j.infbeh.2019.101345",

language = "English",

volume = "57",

journal = "Infant Behavior and Development",

issn = "0163-6383",

publisher = "Elsevier Ltd",

}

TY - JOUR

T1 - A step forward

T2 - Bayesian hierarchical modelling as a tool in assessment of individual discrimination performance

AU - de Klerk, Maartje

AU - Veen, Duco

AU - Wijnen, Frank

AU - de Bree, Elise

PY - 2019/11/1

Y1 - 2019/11/1

N2 - Individual assessment of infants’ speech discrimination is of great value for studies of language development that seek to relate early and later skills, as well as for clinical work. The present study explored the applicability of the hybrid visual fixation paradigm (Houston et al., 2007) and the associated statistical analysis approach to assess individual discrimination of a native vowel contrast, /aː/ - /eː/, in Dutch 6 to 10-month-old infants. Houston et al. found that 80% (8/10) of the 9-month-old infants successfully discriminated the contrast between pseudowords boodup - seepug. Using the same approach, we found that 12% (14/117) of the infants in our sample discriminated the highly salient /aː/ -/eː/ contrast. This percentage was reduced to 3% (3/117) when we corrected for multiple testing. Bayesian hierarchical modeling indicated that 50% of the infants showed evidence of discrimination. Advantages of Bayesian hierarchical modeling are that 1) there is no need for a correction for multiple testing and 2) better estimates at the individual level are obtained. Thus, individual speech discrimination can be more accurately assessed using state of the art statistical approaches.

AB - Individual assessment of infants’ speech discrimination is of great value for studies of language development that seek to relate early and later skills, as well as for clinical work. The present study explored the applicability of the hybrid visual fixation paradigm (Houston et al., 2007) and the associated statistical analysis approach to assess individual discrimination of a native vowel contrast, /aː/ - /eː/, in Dutch 6 to 10-month-old infants. Houston et al. found that 80% (8/10) of the 9-month-old infants successfully discriminated the contrast between pseudowords boodup - seepug. Using the same approach, we found that 12% (14/117) of the infants in our sample discriminated the highly salient /aː/ -/eː/ contrast. This percentage was reduced to 3% (3/117) when we corrected for multiple testing. Bayesian hierarchical modeling indicated that 50% of the infants showed evidence of discrimination. Advantages of Bayesian hierarchical modeling are that 1) there is no need for a correction for multiple testing and 2) better estimates at the individual level are obtained. Thus, individual speech discrimination can be more accurately assessed using state of the art statistical approaches.

KW - Autoregressive error structure

KW - Bayesian hierarchical modeling

KW - Hybrid visual fixation

KW - Individual analysis

KW - Speech sound discrimination

UR - http://www.scopus.com/inward/record.url?scp=85070693790&partnerID=8YFLogxK

U2 - 10.1016/j.infbeh.2019.101345

DO - 10.1016/j.infbeh.2019.101345

M3 - Article

AN - SCOPUS:85070693790

SN - 0163-6383

VL - 57

JO - Infant Behavior and Development

JF - Infant Behavior and Development

M1 - 101345

ER -

A step forward: Bayesian hierarchical modelling as a tool in assessment of individual discrimination performance

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this