Questionable research practices in applied linguistics: an examination of frequency, severity, and AI-related issues.

The concept of questionable research practices (QRPs) has recently emerged as a significant concern in the field of applied linguistics, a discipline that has traditionally distinguished between macro- and micro-ethics. Despite the critical need for studying QRPs, they have been somewhat overlooked within the applied linguistics community. This mixed-methods study aims to shed light on the prevalence and severity of QRPs among Iranian applied linguists, with a particular focus on issues related to the use of artificial intelligence (AI) tools. A total of 160 Iranian applied linguists participated in the study, completing a 48-item questionnaire adapted from Larsson et al. (2023). Fifteen participants also took part in follow-up scenario-based interviews concerning AI-related QRPs. The most common QRPs identified were the selection of variables for convenience and P-hacking, while the most severe were plagiarism and authorship issues. Our qualitative data analysis led to the emergence of four themes: lack of AI related regulations, excessive use of AI as a threat, AI disclosure dilemma, and Al deficiencies. The study concludes that the primary drivers of QRPs among Iranian applied linguists are a lack of systematic research ethics training, insufficient statistical literacy, and an absence of AI regulations.