Yes, it is methodologically possible for a factor in exploratory factor analysis (EFA) to include items linked to two theoretical frameworks, especially if the items share a conceptual foundation, such as perceptions of intelligence. However, this may indicate overlapping constructs or potential measurement issues if the theoretical distinction between fixed and growth mindsets is strong. To address this, you could proceed with a first-order confirmatory factor analysis (CFA) by grouping items into subscales for fixed and growth mindsets, allowing for a total score while maintaining theoretical distinctions. Alternatively, a bifactor CFA could model a general perception of intelligence factor alongside specific fixed and growth mindset factors. A second-order CFA could be considered if the first-order factors (fixed and growth mindsets) emerge distinctly. It’s also important to revisit item design to ensure alignment with theoretical constructs and use robust statistical methods to confirm the factor structure. This approach allows for a balance between empirical findings and theoretical underpinnings.

Of course. A factor structure is ultimately a variances/covariances structure on which theoretically constructed labels are placed. And several theoretical models (i. e. set of labels) can present a similar factor structure. The EFA indicates whether the data fit a model and the FCA indicates whether the model fits to the data. The combined use of both is often a recommended strategy.

Dear Samuel Mores Geddam and José Luis Palacios thank you for your answers.

Samuel Mores Geddam "To address this, you could proceed with a first-order confirmatory factor analysis (CFA) by grouping items into subscales for fixed and growth mindsets, allowing for a total score while maintaining theoretical distinctions." This is how I try to proceed. I did it on AMOS and fit is perfect. But I have to find some references to defend this point of view in front of thesis monitoring committee. I read Kline, R. B., Tabachnick, B. G., & Fidell, L. S. and Byrne, B. M. but couldn't find a specific sentence. But I didn't try a bi-factor model. I read about bi-factor but I couldn't figure out how to apply this for my case.

José Luis Palacios "The combined use of both is often a recommended strategy." This is a recommended way for using EFA and CFA, but I need a reference.

There´s a lot of possible references for combined use of EFA and CFA, but let me suggest Herbert W. Marsh, Alexandre J.S. Morin, Philip D. Parker, and Gurvinder Kaur (2014). Structural Equation Modeling: An Integration of the Best Features of Exploratory and Confirmatory Factor Analysis, Annual Review of Clinical Psychology, 10: 85-110. (doi: 10.1146/annurev-clinpsy-032813-153700)

José Luis Palacios I found the paper and took the necessary notes.

Then, it can be said that if the fit indexes (construct validity, etc.) are fine, a factor containing different theoritical concepts can be modified according to the postulated theory.

Yes, indeed. Different theories can give different results, but the same theory can also be expressed by the same or different factors. EFA can show a structure in “numerical” terms, and the researcher is the one who names the factors of the structure. The same can be said of CFA, mutatis mutandis, of course. So, be careful: numbers do not speak for themselves.