Hans-Friedrich Köhn

Section 1

Program Area: Quantitative

Assistant Professor of Psychology

User Photo

Contact Information:

  • Address:
    425 Psychology Bldg.
    603 E. Daniel Street
    M/C 716
    Champaign, IL 61820

Research Description

My current research concerns three lines of work:

(1) Combinatorial data analysis of individual differences based on multiple proximity matrices observed from different data sources (e.g., subjects, experimental conditions, time points);

(2) Large-scale nonmodel-based clustering, with particular focus on the p-median model;

(3) Cognitively Diagnostic Modeling



Selected Publications:

Köhn, H.-F. (in press). Citation classics commentary on Greenhouse, S. W., & Geisser, S. (1959). On methods in the analysis of profile data.  Psychometrika.

Köhn, H.-F., & Chiu, C.-Y. (in press). A procedure for assessing the completenessof the Q-matrices of cognitively diagnostic tests.  Psychometrika.

Köhn, H.-F., & Chiu, C.-Y.  (2016). A proof of the duality of the DINA model and the DINO model.  Journal of Classification, 33, 171-184.

Chiu, C.-Y., & Köhn, H.-F.   (2016).The Reduced RUM as a logit model: Parameterization and constraints.  Psychometrika, 81, 350-370.

Chiu, C.-Y., & Köhn, H.-F.   (2016).  Consistency of cluster analysis for cognitive diagnosis: The Reduced Reparameterized Unified Model and the General Diagnostic Model.  Psychometrika, 81, 585-610.

Köhn, H.-F., & Hubert, L. J. (2015).  Hierarchical cluster analysis.  Wiley StatsRef: Statistics Reference Online (WSR).

Köhn, H.-F., Chiu, C.-Y., & Brusco, M. J. (2015).  Heuristic cognitive diagnosis when the Q-matrix is unknown.  British Journal of Mathematical and Statistical Psychology, 68, 268-291.

Köhn, H.-F. (2011).  A review of multiobjective programming and its application in quantitative psychology.  Journal of Mathematical Psychology, 55, 386-396. 

Köhn, H.-F. (2010).  Representation of individual differences in rectangular proximity data through anti-Q matrix decomposition.  Computational Statistics and Data Analysis, 54, 2343-2357.

Köhn, H.-F., Steinley, D., and Brusco, M.J. (2010).  The p-median model as a tool for clustering psychological data.  Psychological Methods, 15, 87-95.

Brusco, M.J., and Köhn, H.-F. (2009).  Clustering qualitative data based on binary equivalence relations: a variable neighborhood search procedure for the clique partitioning problem.  Psychometrika, 74, 685-703.

Brusco, M. J., and Köhn, H.-F. (2009).  Exemplar-based clustering via simulated annealing: a comparison to affinity propagation and vertex substitution. Psychometrika, 74, 457-475.

Brusco, M.J., and Köhn, H.-F. (2008).  Optimal partitioning of a data set based on the p-median model.  Psychometrika, 73, 89-105.

Brusco, M.J., and Köhn, H.-F. (2008).  Comment on “Clustering by passing messages between data points”.  Science, 319, 726c.

Brusco, M.J., Köhn, H.-F., and Stahl, S. (2008).  Heuristic implementation of dynamic programming for matrix permutation problems in combinatorial data analysis. Psychometrika,  73, 503-522.

Section 2


Facilities Information

Building Remodeling Projects

Click on the title to check out these pictures from the classroom remodeling project in the basement. A second project to replace the elevators will begin in September. Beginning with the freight elevator, each elevator will be out of service for three months. To get quickly from one floor to another, and improve your fitness, we will encourage the use of the stairs. The repainting of the northeast stairwell has been completed, and the painter is starting on the southwest stairwell.