I have not calculated Cronbach’s Alpha. Because of that, I cannot interpret a reliability coefficient at this stage.
Still, I feel moderately confident about my scale’s p...
I have not calculated Cronbach’s Alpha. Because of that, I cannot interpret a reliability coefficient at this stage.
Still, I feel moderately confident about my scale’s potential reliability because I tried to write items that are clearly connected to the same main constructs: Human–AI collaboration and self-regulated learning in academic writing.
To promote internal consistency, I used items that focus on related behaviors, such as idea generation, revision, monitoring, evaluation, and critical use of AI. I also tried to keep the wording simple, specific, and consistent in tone so respondents would interpret the items in similar ways. In addition, most items were written as single-idea statements rather than combining multiple ideas in one question.
One challenge I faced was making sure the items were similar enough to measure the same construct, but not so repetitive that they sounded identical. Another difficulty was avoiding overlap between AI use and self-regulated learning, because these two areas are closely related in my topic. In the next draft, I would improve the scale by checking whether some items are redundant, revising any vague wording, and piloting the questionnaire with real responses before calculating Cronbach’s Alpha.