From when the black box of authorial identity has been unpacked, the paucity of the authorial identity model based on a comprehensive theoretical framework addressed the need to establish a robust one (Cheung, Stupple, & Elander, 2015). The current study was comprised of three main phases including hypothesizing a model of authorial identity, developing and validating a questionnaire based on the model, and finally testing the model based on the questionnaire data. The participants, including M.A. and Ph.D. students, were 30 for initial piloting, 60 for reliability estimation, 140 for exploratory factor analysis, and 175 for confirmatory factor analysis. At first, drawing on Ivanič’s (1998) model of writer identity and Prior’s (2001) ways of classifying voice, reviewing the related literature, and consulting with a cadre of experts, a model of authorial identity was proposed. Secondly, a questionnaire was developed and validated based on the hypothesized model. The reliability of the questionnaire, estimated through Cronbach’s alpha, was 0.73. Following that, exploratory factor analysis identified four components, namely authorial voice and identity, authorial persona, authorial background, and authorial style. Ultimately, SEM was run using AMOS in the confirmatory factor analysis phase to test the model. The results of this multi-phase research are presented and discussed for underlining the key role of authorial identity in academic writing for both novice and professional academicians.
