H. de Vries

Ethology & Socio-ecology Group, Department of Comparative Physiology, Utrecht University, Utrecht, The Netherlands

At least three topics are important in the analysis of social dominance relationships between the individuals living in a social group:

  1. The linearity test. Do the individuals form a linear dominance hierarchy? That is: how can one test statistically that the linearity in the set of dominance relationships is significantly strong?
  2. The linear ordering problem. If the linearity is significant, how can the individuals be ordered into a (nearly) linear rank order?
  3. The unidirectionality test. Is the exchange of dominance interactions among each pair of individuals sufficiently unidirectional?

I will present a recently developed test of linearity in a set of dominance relationships that contains unknown or tied relationships. Next I will briefly discuss an algorithm for finding the (near) optimal ranking of individuals based on two criteria: minimization of the number of reversals and minimization of the total strength of the reversals. If time allows I will discuss a statistical test of unidirectionality. These three methods are part of the newest version of MatMan, a program for the analysis of sociometric matrices and behavioural transition matrices. A previous version of MatMan is described in Behaviour (125:157-175, 1993).

Paper presented at Measuring Behavior '96, International Workshop on Methods and Techniques in Behavioral Research, 16-18 October 1996, Utrecht, The Netherlands