Going in the Right Direction: A Tutorial to Directional Hypothesis Testing Using the BFpack Module in JASP

Many scientific theories predict not just whether an effect exists, but also how multiple parameters relate—for example, that one mean exceeds another or that several effects follow a specific order. Such expectations can be formulated as directional hypotheses with equality and order constraints. Yet, researchers often test these indirectly through post-hoc analyses of individual parameters, which (1) provide only indirect evidence for the theory, (2) require multiplicity corrections that lower power, and (3) can lead to selective emphasis on statistically significant results. The goal of the current tutorial paper is two-fold. First, it introduces the BFpack module in JASP for testing directional hypotheses in a direct manner using Bayes factors. BFpack offers default Bayesian procedures for common designs—$t$ tests, (M)AN(C)OVA, regression, and multi-group correlations—without the need for manual prior specification. By evaluating theory-based hypotheses jointly, the method avoids multiplicity, increases power, and discourages selective reporting.Through three worked examples, we demonstrate how BFpack simplifies directional hypothesis testing in practice and helps researchers move beyond traditional post-hoc testing toward theory-driven inference.