Department of Statistics
London School of Economics
Office: King's Chambers 2.08
Research interestsLatent variable modelling (factor analysis, structural equation modeling, item response theory, latent class models)
Longitudinal data analysis
Analysis of ordinal and ranking data
Applications to Social Sciences, particularly to Marketing (Market Research, Consumer Behavior Analysis)
Publications and Preprints
- Kuha, J., Katsikatsou, M., and Moustaki, I. (forthcoming) Latent variable modelling with non-ignorable item non-response: A general framework and multigroup models for cross-national analysis. Journal of the Royal Statistical Society - Series A. Accepted for publication. [Preprint]
- Kuha, J., Butt, S., Katsikatsou, M., and Skinner, C. J. (2017). The effect of probing “Don't Know” responses on measurement quality and nonresponse in surveys. To appear in Journal of the American Statistical Association. http://dx.doi.org/10.1080/01621459.2017.1323640
- Katsikatsou, M. and Moustaki, I. (2016) Pairwise likelihood ratio tests and model selection criteria for structural equation models with ordinal variables. Psychometrika, 81 (4), 1046-1068.
- Katsikatsou, M., Moustaki, I., Yang-Wallentin, F., and Jöreskog, K. G. (2012) Pairwise likelihood estimation for factor analysis models with ordinal data. Computational Statistics and Data Analysis, 56, 4243-4258.
- Katsikatsou, M. (2013). Composite Likelihood Estimation for Latent Variable Models with Ordinal and Continuous, or Ranking Variables. Uppsala University, Sweden. PhD Thesis. [PDF]
- Katsikatsou, M., and Moustaki, I. Pairwise likelihood estimation for confirmatory factor analysis models with ordinal variables and data that are missing at random. Journal of the Royal Statistical Society - Series C. Submitted - Under Review.
- Limited information goodness-of-fit tests for pairwise likelihood estimation under Simple Random Sampling and complex survey sampling (working title). With Moustaki, I. and Skinner, C.
- Katsikatsou, M. (2017) The Pairwise Likelihood Method for Structural Equation Modelling with ordinal
variables and data with missing values using the R package lavaan.
To replicate the analyses presented in the tutorial download the Data. The R commands for the analyses, also provided in the tutorial document, are gathered in the script file R_commands.
- Katsikatsou, M. and Kuha, J. (2016) Latent variable modelling of cross-national survey data. Published on ESS EduNet website.
SoftwareR package lavaan: Code contributor for: estimator="PML", i.e. contributed the code for
- the pairwise likelihood estimation of structural equation models with ordinal variables,
- the pairwise likelihood ratio test for the overall fit of a model and for nested models),
- the model selection criteria PL-AIC and PL-BIC, and
- the PML adaptations in the case of missing data such as complete-pairs PML, available-case PML, and doubly-robust PML.
My research on Google Scholar.