Improving the touchscreen-based food approach-avoidance task: remediated block-order effects and initial findings regarding validity

Approach biases to foods may explain why food consumption often diverges from deliberate dietary intentions. When cognitive resources are depleted, implicit responses may contribute to overeating and overweight. Yet, the assessment of behavioural biases with the approach-avoidance tasks (AAT) is often unreliable. We previously addressed methodological limitations of the AAT by employing naturalistic approach and avoidance movements on a touchscreen (hand-AAT) and instructing participants to respond based on the food/non-food distinction. In the consistent block, participants were instructed to approach food and avoid objects while in the inconsistent block, participants were instructed to avoid foods and approach objects. Biases were highly reliable but affected by the order in which participants received the two task blocks. In the current study, we aimed to resolve the block order effects by increasing the number of blocks from two to six and validate the hand-AAT with the implicit association task (IAT) and self-reported eating behaviours. We replicated the presence of reliable approach biases to foods and further showed that these were not affected by block order. Evidence for validity was mixed: biases correlated positively with external eating, food craving and aggregated image valence ratings but not with within-participants differences in desire to eat ratings of the images or the IAT. We conclude that hand-AAT can reliably assess approach biases to foods that are relevant to self-reported eating patterns and were not probably confounded by block-order effects.