How the HPEI was validated
The validity of the structure (or construction) of the Enneagram was tested using principal component factor analysis methods with varimax rotation. The analyses enable us to observe that there are indeed 9 distinct factors corresponding to the 9 Enneagram Types. Each item therefore indeed belongs to one, single factor. The percent of variance explained by the 9 factors is 55.34% which is satisfactory.
The consistency or internal reliability is satisfactory for the 9 Types. This is evaluated using Cronbach’s Alpha coefficient which enables assessment of the tool’s capacity to reliably measure the target construct while minimising measurement errors. A scale is considered reliable if the coefficient is greater than 0.7. Therefore this means that the HPEI measures the Enneagram Types in a reliable way. The coefficients of variation are also satisfactory, indicating a good dispersion of the responses around the mean.
Moreover it is observed that the score distribution on each of the 9 HPEI dimensions is close to normal.
For 8 of the 9 Types in the HPEI (except 3 which according to the initial samples characterise men better than women) the dimensions do not vary depending on age or sex. This means it is not necessary to construct different standardisations according to sex and age groups.
The convergent/discriminant validity tests with the Big Five, which is the international benchmark for personality questionnaires in the academic world, produce significant results. The crossover between the 5 BigFive variables (agreeableness, conscientiousness, emotional stability, extraversion and openness) produces correlations from .0 to .58. This means that the HPEI, while having significant convergences with the Big Five, indeed measures "something other" than the latter and so provides discriminant information compared with this tool. The same applies when compared with the MBTI.
Between the Big Five and the HPEI, a remarkably significant correlation is observed between for example: Type 1 and conscientiousness, Type 2 / 9 and agreeability, Type 3 / 7 / 8 and extroversion..
Hierarchical multiple regression analyses allow us to verify whether a tool enables us to "predict" certain behaviours, focuses of attention or personality variables. We conducted these analyses in relation to educational guidance (Schein career anchors) and team roles (MTRi). Compared with both the Big Five and the MBTI, the HPEI proves to be significantly more predictive than the other tools.