Digman (1997) performed higher order factor analysis on several Big 5 intercorrelation matrices and proposed that the Big 5 could be partially explained in terms of two higher order factors: Alpha (Agreeableness, Conscientiousness, Emotional Stability) and Beta (Extraversion, Openness).
Musek (2007) presented evidence for a Big 1 factor of personality. Musek and others observe that while the Big 5 factor structure was based on orthogonal rotations, Big 5 scales based on items tend to intercorrelate. With regards to interpretation, Musek states that "[it seems plausible therefore to assume that, beyond the connections with semantic factors and social desirability, The Big One is substantially related to the basic dimensions of emotionality (affect), personal well-being, and self-esteem."
To empirically test the Big 1, Musek presented results from three datasets. Results showed that (a) the first factor of a factor analysis of the Big 5 explained approximately 40 to 50 percent of the variance (b) a reasonable amount of variance in test items was explained by the first factor, (c) several confirmatory factor analytic models indicating reasonable fit for a hierarchical model of the Big 5 that includes the two higher order factors proposed by Digman subsumed under a global personality factor.
van der Linden et al 2010 extend the work of Musek (2007) by presenting a meta-analysis of Big 5 intercorrelation matrices
in order to assess the evidence for a General Factor of Personality.
Table 2 shows the meta-analytic average scale intercorrelations for the big 5 measures included in the meta-analysis. It shows how across a wide range of measures and measurement procedures, the Big 5 scales moderately intercorrelate. The average absolute unadjusted correlations between the Big Five was 0.23.
They then show how across a range of Big 5 measures, measurement procedures, and samples that a single factor accounts for between 42 and 78 percent of variance in the Big 5 scale scores.
With regards to interpretation the van der Linden et al present various views on what the factor represents including views that see it as a statistical or measurement artefact and views that see it as a meaningful construct. One interpretation is that it is a general goodness of character factor or perhaps a social adjustment factor.
The authors conclude with the following:
The existence of a higher-order GFP does not invalidate the clinical,
vocational, or theoretical importance of lower-order factors. It is an
empirical and practical question as to which level provides the best
predictor for a given criterion. Since the personality facets that
exist below the Big Five factors lie closest to the behavior
expressed, they are often more diagnostic or better predictors than
higher order traits (Sackett & Lievens, 2008).
Concluding thoughts: It seems that there is a meaningful higher level factor. While it may reflect social desirability, such a characteristic may reflect as much true variance in social desirability as it does a response style. As an aside, I have noticed that in experiments where people are asked to persent themselves in a positive way for a job, the big 5 intercorrelations increase dramatically. In broad terms the global factor of personality also reflects broadly the sign of meta-analytic correlations of Big 5 with job performance (i.e., positive correlations for extraversion, openness, emotional stability, etc.).
I have pondered to what extent measures of the Big 5 are unnecessarily loaded to see various poles as socially desirable. For example, a quick look at this IPIP measure of the Big 5 shows quite clearly how positive aspects of for example introversion are lacking (e.g., the item is "I have little to say" rather than something like "I enjoy my own company") and negative aspects of extraversion are lacking (e.g., the item is "I am the life of the party" rather than "I like to dominate social situations").
I'd also like to see more item-level analyses and the relative support for the Global personality factor.
Clearly, a lot of detail is lost in the reduction from the Big 5 to the Big 1, but it also seems to reflect a useful summary of personality for some predictive purposes.
- Digman, J. M. (1997). Higher-order factors of the Big Five. Journal of Personality and Social Psychology, 73, 1246–1256.
- Musek, J. (2007). A general factor of personality: Evidence for the Big One in the
Five-Factor Model. Journal of Research in Personality, 41, 1213–1233.
- van der Linden, D., te Nijenhuis, J., & Bakker, A. B. (2010). The general factor of personality: A meta-analysis of Big Five intercorrelations and a criterion-related validity study. Journal of Research in Personality, 44(3), 315-327. PDF
- Sackett, P. R., & Lievens, F. (2008). Personnel selection. Annual Review of Psychology,