Abstract
A problem that tends to be ignored in the statistical analysis of experimental data in the language sciences is that responses often constitute time series, which raises the problem of autocorrelated errors. If the errors indeed show autocorrelational structure, evaluation of the significance of predictors in the model becomes problematic due to potential anti-conservatism of p-values.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The parametric coefficients suggest that regularity is irrelevant as predictor of naming times, that singulars are named faster than plurals, that words with voiced initial segments have longer naming times, as do words with a large number of words at Hamming distance 1 at the initial segment. Words with a greater Shannon entropy calculated over the probability distribution of their inflectional variants elicited shorter response times. A thin plate regression spline for log-transformed word frequency suggests a roughly U-shaped effect (not shown) for this predictor.
- 2.
For this to work properly, it is necessary to use treatment contrasts for ordinal factors, in R: options(contrasts = c("contr.treatment", "contr.treatment")).
- 3.
The details of the coefficients in the present model differ from those obtained in the analysis of Baayen [1]. Thanks to the factor smooths for subject and compound and the inclusion of a thin plate regression spline for word frequency, the present model provides a better fit (aic 177077.4 versus 187308), suggesting the present reanalysis may provide a more accurate window on sex-specific realizations of compounds’ pitch.
- 4.
Data points with an absolute amplitude exceeding 15 μV, approximately 2.6% of the data points, were removed to obtain an approximately Gaussian response variable.
References
Baayen RH (2013) Multivariate statistics. In: Podesva R, Sharma D (eds) Research methods in linguistics. Cambridge University Press, Cambridge, pp 337–372
Baayen RH, Milin P (2010) Analyzing reaction times. Int J Psychol Res 3:12–28
Baayen R, Vasishth S, Bates D, Kliegl R (2015) Out of the cage of shadows. arxiv.org. http://arxiv.org/abs/1511.03120
Bates D, Mächler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models using lme4. J Stat Softw 67(1):1–48
Broadbent D (1971) Decision and stress. Academic Press, New York
DeCat C, Baayen RH, Klepousniotou E (2014) Electrophysiological correlates of noun-noun compound processing by non-native speakers of English. In: Proceedings of the first workshop on computational approaches to compound analysis (ComAComA 2014). Association for Computational Linguistics and Dublin City University, Dublin, Ireland, pp 41–52
DeCat C, Klepousniotou E, Baayen RH (2015) Representational deficit or processing effect? A neuro-psychological study of noun-noun compound processing by very advanced l2 speakers of English. Front Psychol (Lang Sci) 6:77
De Vaan L, Schreuder R, Baayen RH (2007) Regular morphologically complex neologisms leave detectable traces in the mental lexicon. Ment Lexicon 2:1–23
Koesling K, Kunter G, Baayen RH, Plag I (2012) Prominence in triconstituent compounds: pitch contours and linguistic theory. Lang Speech 56(4):529–554
Lin X, Zhang D (1999) Inference in generalized additive mixed models using smoothing splines. J R Stat Soc Ser B 61:381–400
Paeschke A, Kienast M, Sendlmeier W (1999) F0-contours in emotional speech. In: Proceedings of the 14th International Congress of Phonetic Sciences, vol 2, pp 929–932
Sanders A (1998) Elements of human performance: reaction processes and attention in human skill. Lawrence Erlbaum, Mahwah, NJ
Tabak W (2010) Semantics and (ir)regular inflection in morphological processing. PhD thesis, University of Nijmegen. Ponsen & Looijen, Ede
Taylor TE, Lupker SJ (2001) Sequential effects in naming: a time-criterion account. J Exp Psychol Learn Mem Cogn 27:117–138.
Traunmüller H, Eriksson A (1995) The frequency range of the voice fundamental in the speech of male and female adults. Institutionen för lingvistik, Stockholms Universitet, S-106 91 Stockholm, Sweden
Trouvain J, Barry WJ (2000) The prosody of excitement in horse race commentaries. In: ISCA tutorial and research workshop (ITRW) on speech and emotion
Welford A (1980) Choice reaction time: basic concepts. In: Welford A (ed) Reaction times. Accademic Press, New York, pp 73–128
Wilkinson G, Rogers C (1973) Symbolic description of factorial models for analysis of variance. Appl Stat 22:392–399
Wood SN (2006) Generalized additive models. Chapman & Hall/CRC, New York
Wood SN (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models. J R Stat Soc (B) 73:3–36
Wood SN (2013) On p-values for smooth components of an extended generalized additive model. Biometrika 100:221–228
Wood SN (2013) A simple test for random effects in regression models. Biometrika 100:1005–1010
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this chapter
Cite this chapter
Baayen, R.H., van Rij, J., de Cat, C., Wood, S. (2018). Autocorrelated Errors in Experimental Data in the Language Sciences: Some Solutions Offered by Generalized Additive Mixed Models. In: Speelman, D., Heylen, K., Geeraerts, D. (eds) Mixed-Effects Regression Models in Linguistics. Quantitative Methods in the Humanities and Social Sciences. Springer, Cham. https://doi.org/10.1007/978-3-319-69830-4_4
Download citation
DOI: https://doi.org/10.1007/978-3-319-69830-4_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-69828-1
Online ISBN: 978-3-319-69830-4
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)