Stability of AI-attitude profiles over time

Establish whether the attitudinal profiles concerning AI use among PhD researchers remain stable over time.

Background

The analysis uses cross-sectional data from a self-selected international sample of PhD researchers. This design captures attitudes at one point in time but cannot determine whether the four latent classes—“division of labour,” “status quo,” “all-purpose,” and “undecided”—represent durable orientations or temporary responses to changing familiarity, access, institutional expectations, and peer practices.

The authors therefore identify longitudinal research following the same researchers over time as necessary for determining how these attitudes evolve and whether the observed profiles are stable.