Hi Apoora,

what was the sample? If you have a focused sample (e.g., students) that is related to C or O, than endogeneous selection can occurr with substantially changed relationships.

Here is a blog article that illustrates this:

http://www.the100.ci/2017/03/14/that-one-weird-third-variable-problem-nobody-ever-mentions-conditioning-on-a-collider/

Beyond that, the identification/selection of measures can affect the correlation. While the big five are often perceived as latent factors, there measures/scales are extremely heterogeneous making each of the big fives rather a broad category of facets which "count" as, e.g., conscientousness. Depending on the selection, openess-related facets could be prevalent.

HTH

Holger

Yes.

Holger Steinmetz Thank you so much for the link. I'll go through it.

My sample is not focused, its open for general consumers. I have not found any literature support for correlation of both the traits. I found the respondents in my sample (though i am still doing pilot) to be giving same answers to the questions of both the constructs.

Matthew Kerry If you could please share some research evidence to elaborate your answer.

Thank you so much.

Hi,

actually I don't have any meta-analytical correlation between C and O--hence I do not know how's the correlation between both. I would actually guess that it is positive.

One final remark, though, is that correlations can chance due to confounding. If your sample consists of, say, two subgroups with different means in both C and O, correlation without considering these differences can change the correlation or even change the direction. This has become popular as the Simpson's Paradox, see

Hernán, M. A., Clayton, D., & Keiding, N. (2011). The Simpson's paradox unraveled. International journal of epidemiology, 40(3), 780-785.

Best,

Holger

Referring to this questions I recommend you to read:

McCrae, R. R., & Costa, P. C. ( 1997). Conceptions and correlates of openness to experience. In Hogan R., Johnson J., & Briggs S. ( Eds.), Handbook of personality psychology. San Diego. Academic Press

Best regards,

Thank you so much Marius Babici and Holger Steinmetz ... These references have been very helpful.

Warm regards,

Apoorva, you may also consider the problem of method effects in your data.

There is a nice article showing that the Big Five factors were correlated when self-reports were used, but they were uncorrelated, when different methods were used.

Biesanz, J. C., & West, S. G. (2004). Towards understanding assessments of the Big Five: Multitrait-multimethod analyses of convergent and discriminant validity across measurement occasion and type of observer. Journal of Personality, 72, 845-876.

Article Towards Understanding Assessments of the Big Five: Multitrai...

ABSTRACT

Multitrait-multimethod analyses were used to examine the degree of convergent and discriminant validity of the Big Five. Phase 1 examined self-reports of the Big Five across three measurement occasions. Self-reports of the Big Five traits were stable, but were moderately intercorrelated. Phase 2 examined assessments of the Big Five across different types of informants (self, peer, and parent). Assessments converged across types of informants and, importantly, there was no evidence of correlation between the Big Five traits across the perspectives of different types of informants. The present results suggest that the degree of orthogonality of the Big Five traits depends on the source of the data. A single informant produces Big Five traits that are intercorrelated, whereas diverse informants tend to produce a much more orthogonal structure. Discussion focuses on methodological considerations in examining levels of convergent and discriminant validity and the theoretical implications.

Hope this provides a different perspective on this problem.

Kind regards, Karin