Post-hoc hypothesizing (HARKing) presents data-driven results as a priori predictions, which fundamentally compromises the validity of traditional statistical testing and contributes to the replication crisis.
The retrieved literature (such as Kerr 1998 and Rastogi et al. 2021) consistently identifies HARKing (Hypothesizing After the Results are Known) as a detrimental practice that invalidates traditional statistical testing by masking exploratory findings as confirmatory hypotheses, leading to inflated type I error rates and non-reproducible results. No papers refute this claim.