Measuring Behavior 2005: Jähkel et al.

Principal components factor analysis of different behavioral and neurochemical items in mice: A method to reveal contextual relations of behaviors and neurochemistry

M. Jähkel, L. Schiller and J. Oehler

Department of Psychiatry and Psychotherapy, Technical University, Dresden, Germany

A variety of equipments allows more and more detailed observations and descriptions of animal’s behavior. Commonly a number of parameters is measured ranging from the frequency and duration of items to the intensity of a behavior scored by means of a graded system. But the increasing number of parameters is difficult to handle especially when we need to choose the relevant parameters to be analyzed further. The principal components factor analysis is an effective method to reveal interrelationships between measured items assembling parameters into factor groups by means of a multicorrelative procedure.

The study reports mice data collected in the following behavioral tests: magnetic impulse based running wheel test measuring 90o movements of the wheels; light beam based open field and plus maze tests measuring place and time characteristics of movement sequences; video recorded social intruder test with subsequent manual based evaluations measuring frequency and duration of different social and neutral behaviors. In addition, neurochemical items are reported which were obtained immediately after the social intruder tests.

All parameters collected are subjected to test-related principal components factor analyses. Important steps and appropriate results (K-M-O quality criteria, communality, anti-image matrix, rotated factor matrix, factor loadings) are described for each test performed. Furthermore the result of an overall factor analysis including the most important behavioral and neurochemical data is presented. A functionally relevant differentiation into drive, curious eager and irritability is revealed regarding activity parameters directed to explore an environment. Behavioral items measured during the social intruder test are to differentiate into social af.ne against social aversive behaviors but additionally drive, curious eager and irritability as relevant behavioural axes are also found. Finally, exemplary behavioral measurements are shown to demonstrate housing and sex effects on mice’s behavior.

Interrelationships between the revealed behavioural axes and neurochemistry as well as the power of the multicorrelative principal component factor analysis are discussed.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.