==============================================================================
LOG-DAY MODEL + COHEN'S f + POWER -- naive_a2_d0_10   [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=3, control n=7   nrep=120, alpha=0.05

--- fitted on real data ---
stim x log(day) interaction: F(1,98)=6.255  p(resid)=0.01404  p(Satt)=0.01413 (df=93)
  honest per-animal random slope (log-day): F(1,6.5)=3.64  p=0.1012
Cohen's f (interaction, partial eta^2=0.019) = 0.138  (small-medium; f: .10 small, .25 medium, .40 large)

--- power simulation (log-day ground truth: stim:logday=+0.07796, ratSD=0.04949, resSD=0.1018) ---

 true stim:log(day) = +0.07796 (100% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.25            |  0.57  <- observed
  5        |  0.71            |  0.84
  8        |  0.92            |  0.95
  12       |  1.00            |  1.00
  16       |  1.00            |  1.00
  24       |  1.00            |  1.00

 true stim:log(day) = +0.03898 (50% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.09            |  0.15  <- observed
  5        |  0.14            |  0.26
  8        |  0.41            |  0.46
  12       |  0.52            |  0.56
  16       |  0.78            |  0.78
  24       |  0.88            |  0.89

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