Based on the factor loadings, a CFA model with 3 content and 6 method factors was completed.
The substantive meaning of the general factor and the method factor is further discussed based on their correlations with a series of personality and well-being measures.
However, it is known that the solution depends on which method factor is excluded from the model as the reference.
Each method factor was examined separately and then both were modelled simultaneously.
The loadings for each method factor were constrained equal under the assumption that a method effect should exert a relatively equal influence on all items, but also in order to maximize degrees of freedom given the medium sample size.
In response to the comments of Reviewer-2 (Professor Mulaik), we attempted to check whether the progressive methods factors were necessarily independent of the corresponding true NK activity factors, by freeing the corresponding covariances (which were constrained to be equal).
The variances of the two methods factors were constrained to be equal because the procedural similarity in the measurement series initially led us to suspect that routine within-series laboratory variations might propagate proportionally.
Furthermore, only 36% of the factor loadings of the manifest variables on the latent common methods factor were significant at 5% level, not satisfying the convergent validity criteria.
When a latent common methods factor was added to the model, there was a significant improvement in model fit: Δχ 33)= 74.84, p<.01 (applying the scaled difference in χs test for nested models [ 57]), CFI=.98, NNFI=.98, SRMR=.07, and RMSEA=.03.
This is the case for the SW Prony, so for this method, the damping factors are set equal to zero.
