Abstract
In this paper we introduce a modified version of existing dimensionality reduction method for binary data, weighted logistic principal component analysis (WLPCA). We propose to fit the basis vectors of the latent natural parameter subspace in a successive procedure instead of fitting them at ones, so the vectors will be sorted by an explanation power of the data in term of model likelihood. Based on our modified WLPCA model, we present a methodology for analyzing binary (true/false) questionnaires. The purpose of the methodology is to bring the authors of questionnaires a global overview of relationships between questions based on the correlations of binary answers. In the experiment we employ our proposed model to analyze psychiatric questionnaire, namely the Junior Temperament and Character Inventory (JTCI). The results suggest that our methodology can yield interesting relationships between questions and that our modified model is better suited for such an analysis as the existing versions of the logistic principal component analysis model.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Asch, M.: Psychometric properties of a french version of the junior temperament and character inventory. European Child Adolescent Psychiatry 18(3), 144–153 (2009)
Collins, M., Dasgupta, S., Schapire, R.E.: A generalization of principal component analysis to the exponential family. In: Advances in Neural Information Processing Systems. MIT Press, Cambridge (2002)
Fodor, I.: A survey of dimension reduction techniques. Tech. rep., Lawrence Livermore National Laboratory (2002)
Lyoo, I.K., Han, C.H., Lee, S.J., Yune, S.K., Ha, J.H., Chung, S.J., Choi, H., Seo, C.S., Hong, K.-E.M.: The reliability and validity of the junior temperament and character inventory. Comprehensive Psychiatry 45(2), 121–128 (2004)
Paulinyová, M., Tiňová, M., Halama, P., Hradečná, Z., Škodáček, I.: Realiability and validity of the slovak version of JTCI. Psychiatrie (2011)
Schein, A., Saul, L., Ungar, L.: A generalized linear model for principal component analysis of binary data. In: 9th Int. Workshop Artificial Intelligence and Statistics, Key West, FL (January 2003)
Schmeck, K., Goth, K., Poustka, F., Cloninger, R.C.: Reliability and validity of the JTCI. International Journal of Methods in Psychiatric Research 10(4), 172–182 (2001)
Zivkovic, Z.: Layered image model using binary PCA transparency masks. In: British Machine Vision Conference, BMVA (2007)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Mažgut, J., Paulinyová, M., Tiňo, P. (2012). Using Dimensionality Reduction Method for Binary Data to Questionnaire Analysis. In: Kotásek, Z., Bouda, J., Černá, I., Sekanina, L., Vojnar, T., Antoš, D. (eds) Mathematical and Engineering Methods in Computer Science. MEMICS 2011. Lecture Notes in Computer Science, vol 7119. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-25929-6_14
Download citation
DOI: https://doi.org/10.1007/978-3-642-25929-6_14
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-25928-9
Online ISBN: 978-3-642-25929-6
eBook Packages: Computer ScienceComputer Science (R0)