People might feel calmer after walking in the woods not just because of the trees and fresh air, but because they expect to feel better—like a mind-over-matter effect.

Participants are randomly assigned to either a real forest environment or a simulated forest environment (e.g., virtual reality with forest visuals and sounds) while being told both are 'real forest bathing'. Cortisol levels are measured pre- and post-exposure. Psychological expectations are measured via validated scales before intervention. This isolates the effect of expectation from environmental exposure.

All participants experience the same real forest environment, but are given different verbal instructions: one group is told 'forest bathing is proven to reduce stress', another is told 'forest bathing has no known effect', and a third receives neutral instructions. Cortisol and self-reported expectations are measured before and after. This tests whether expectation alone alters cortisol response despite identical physical exposure.

Each participant undergoes four conditions: (1) real forest + high expectation, (2) real forest + low expectation, (3) urban setting + high expectation, (4) urban setting + low expectation. Expectation is manipulated via pre-exposure briefing. Cortisol is measured across all conditions. This disentangles the independent and interactive effects of environment and expectation.

Participants self-select into forest bathing or control groups. Expectations about stress reduction are measured before the activity, and cortisol is measured before and after. Statistical models control for baseline cortisol, activity duration, and environmental factors (e.g., noise, temperature) to isolate expectation as a predictor of cortisol change.

In-depth interviews with individuals who report cortisol reductions after forest bathing explore their beliefs about why they felt calmer. Themes are coded for references to psychological expectations, placebo-like beliefs, or environmental factors. This identifies common narratives that may inform quantitative hypotheses.