Computing Eysenck's personality types: a closer look at the standard method

Liyanage (2004) carried out a major study to investigate the relationship between
personality type and choice of English as a Second Language (ESL) learning strategies for
a group of 886 participants in Sri Lanka. Liyanage used the Eysenck Personality
Questionnaire (EPQ) to measure personality type in that study. The manual of the EPQ
(1991) indicates that participants can be assigned to four personality types (Choleric, Sanguine,
Melancholic, and Phlegmatic) by comparing the total scores on two subscales (Extroversion;
Neuroticism). However, in the Sri Lankan study, this total score method resulted in a gross
imbalance in the assignment of participants to specific personality types. The present paper
reports and compares outcomes based on the total score method with an alternative approach to
computing personality type that produced more balanced outcomes. It discusses logical links
between the use of these methods and the outcomes of previous investigations of the links
between personality type and language learning strategies.