Smoothing Spline ANOVA Decomposition of Arbitrary Splines: An Application to Eye Movements in Reading

Hannes Matuschek, Reinhold Kliegl, Matthias Holschneider

Full Text: PDF   Paper Package: MatuschekKlieglHolschneider2015_1.0 tar.gz PID: none yet


The Smoothing Spline ANOVA (SS-ANOVA) requires a specialized construction of basis and penalty terms in order to incorporate prior knowledge about the data to be fitted. Typically, one resorts to the most general approach using tensor product splines. This implies severe constraints on the correlation structure, i.e. the assumption of isotropy of smoothness can not be incorporated in general. This may increase the variance of the spline fit, especially if only a relatively small set of observations are given. In this article, we propose an alternative method that allows to incorporate prior knowledge without the need to construct specialized bases and penalties, allowing the researcher to choose the spline basis and penalty according to the prior knowledge of the observations rather than choosing them according to the
analysis to be done. The two approaches are compared with an artificial example and with analyses of fixation durations during reading.



Baayen R, Kuperman V, Bertram R, Baayen R (2010) Frequency effects in compound processing. Compounding, Amsterdam/Philadelphia: Benjamins : 257–270.

Tremblay A, Baayen RH (2010) Holistic processing of regular four-qord sequences: A behavioral and ERP study of the effects of structure, frequency and probability on immediate free recall. In: Wood D, editor, Perspectives on Formulaic Language: Aquisition and communication, London: The Continuum International Publishing Group. pp. 151–173.

Kosling K, Kunter G, Baayen H, Plag I (2012) Prominence in Triconstituent Compounds: Pitch Contours and Linguistic Theory. Language and Speech .

Kimeldorf G, Wahba G (1970) A correspondence between bayesian estimation on stochastic pro- cesses and smoothing by splines. The Annals of Mathematical Statistics 41: 495–502.

Lin X, Zhang D (1999) Inference in generalized additive mixed modelsby using smoothing splines. Journal of the Royal Statistical Society: .10

Ruppert D, Wand M, Carroll R (2003) Semiparametric regression, volume 12. Cambridge Univ. Pr.

Demidenko E (2004) Mixed models: theory and applications, volume 518. John Wiley & Sons.

Hastie T, Tibshirani R (1987) Generalized additive models: some applications. Journal of the American Statistical Association 82: 371–386.

Gu C (2002) Smoothing spline ANOVA models. Springer Verlag.

Wood SN (2006) Generalized Additive Models. Chapman & Hall, 392 pp.

Wood S (2012). Package ’gamm4’. URL

Gu C (2013). Package ’gss’. URL

R Core Team (2013) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. URL

Wieling M, Nerbonne J, Baayen RH (2011) Quantitative social dialectology: Explaining linguistic variation geographically and socially. PLoS ONE 6.

Wieling M, Bloem J, Nerbonne KM, Timmermeister M, Nerbonne J (2014) Measuring Foreign Accent Strength in English: Validating Levenshtein Distance as a Measure. Language 4: 253 – 269.

Wahba G (1990) Spline models for observational data. Philadelphia: Society for Industrial and Applied Mathematics, 169 pp.

Wood SN (2012). Package ’mgcv’. URL

Kliegl R, Nuthmann A, Engbert R (2006) Tracking the mind during reading: the influence of past, present, and future words on fixation durations. Journal of experimental psychology General 135: 12–35.

Heister J, Würzner K, Kliegl R (2012) Visual Word Recognition. Meaning and context, individuals and development., Sussex, UK: Psychology Press, volume 2, chapter Analysing large datasets of eye movements during reading.

Kliegl R (2007) Toward a perceptual-span theory of distributed processing in reading: A reply to Rayner, Pollatsek, Drieghe, Dlattery, and Reichle (2007). Journal of Experimental Psychology: General 136: 530.

Wotschack C, Kliegl R (2013) Reading strategy modulates parafoveal-on-foveal effects in sentence reading. Quarterly journal of experimental psychology 66: 548–562.