==============================================================================
LOG-DAY MODEL + COHEN'S f + POWER -- prev_f_full   [metric: success RATE]
==============================================================================
model: behavior ~ stim + log(day) + stim:log(day) + (1|rat)   (behavior = success RATE)
log(day) uses 1-indexed training day (our day 0 = paper "Day 1")
observed groups: stim n=12, control n=12   nrep=120, alpha=0.05

--- fitted on real data ---
stim x log(day) interaction: F(1,227)=2.888  p(resid)=0.0906  p(Satt)=0.09072 (df=208)
  honest per-animal random slope (log-day): F(1,23.7)=2.05  p=0.1655
Cohen's f (interaction, partial eta^2=0.006) = 0.079  (small; f: .10 small, .25 medium, .40 large)

--- power simulation (log-day ground truth: stim:logday=+0.02595, ratSD=0.06838, resSD=0.08017) ---

 true stim:log(day) = +0.02595 (100% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.06            |  0.07
  5        |  0.17            |  0.18
  8        |  0.33            |  0.38
  12       |  0.38            |  0.47  <- observed
  16       |  0.55            |  0.62
  24       |  0.64            |  0.60

 true stim:log(day) = +0.01298 (50% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.04            |  0.04
  5        |  0.07            |  0.11
  8        |  0.10            |  0.12
  12       |  0.17            |  0.17  <- observed
  16       |  0.10            |  0.15
  24       |  0.15            |  0.17

Read per-animal as the honest power; LME matches the paper's power code (optimistic).
