==============================================================================
LOG-DAY MODEL + COHEN'S f + POWER -- naive_a2_d6_13   [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=6   nrep=120, alpha=0.05

--- fitted on real data ---
stim x log(day) interaction: F(1,61)=0.040  p(resid)=0.843  p(Satt)=0.843 (df=58)
  honest per-animal random slope (log-day): F(1,7.8)=0.00  p=0.9656
Cohen's f (interaction, partial eta^2=0.000) = 0.010  (small; f: .10 small, .25 medium, .40 large)

--- power simulation (log-day ground truth: stim:logday=+0.01851, ratSD=0.05356, resSD=0.07547) ---

 true stim:log(day) = +0.01851 (100% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.05            |  0.05  <- observed
  5        |  0.09            |  0.08
  8        |  0.05            |  0.03
  12       |  0.04            |  0.06
  16       |  0.07            |  0.08
  24       |  0.03            |  0.07

 true stim:log(day) = +0.009253 (50% of observed)
  N/group  | per-animal power | LME power
  ------------------------------------------
  3        |  0.05            |  0.07  <- observed
  5        |  0.05            |  0.06
  8        |  0.07            |  0.04
  12       |  0.04            |  0.05
  16       |  0.04            |  0.05
  24       |  0.07            |  0.07

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