Expand Methods to describe the three interaction DF treatments (residual,
Satterthwaite, honest random-slope) and why Satterthwaite is inert on the
random-intercept model. Add a 'Honest inference' subsection: the early-phase
interaction reaches honest significance only in the pre-specified pooled-control
scenarios (naive_a2/naive_boxa/mergeNaive d0_5-d0_13, random-slope p 0.014-0.036)
while confounded/small-n windows collapse (mergeA2_full 0.75; unmerge_d0_5 0.13).
Add a 'Scenario notation' legend defining every grouping, window, model family,
p-value column, and the _f (Forouzan) labels.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Every fitlme-based report (lme_*, paper_*, phase_*, and the variations'
analyze.m) now shows, per effect: residual-DF p, Satterthwaite-DF p, and --
for the interaction -- an HONEST test from a per-animal random-SLOPE model
(day|rat), whose Satterthwaite DF collapses toward the animal count.
New: tdcs_random_slope_interaction.m (shared helper). Wired into tdcs_lme,
tdcs_paper_lme, tdcs_phase_lme, variation_analyze; SUMMARY.csv gains
interaction_p_satt / interaction_p_rs. Regenerated all results/, variations/,
matched-effort outputs.
Key point this surfaces: Satterthwaite ~= residual on the random-INTERCEPT
model (the slope's error is at session level), so it does NOT fix
pseudoreplication; the random-slope model does. Effect: full-range mergeA2
interaction 0.009 -> 0.75 (collapses); unmerge_d0_5 0.015 -> 0.13 (n.s.);
the pooled-control early windows survive honestly (naive_a2_d0_5 0.001 ->
0.028; naive_boxa_d0_5 0.003 -> 0.036). Suite 42/42.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add make_crossstudy_compare.m producing crossstudy/result.txt +
crossstudy_summary.csv: current control (A2, A2+Naive) and anodal (B2) arms
vs the previous study's a2_f/b2_f, on per-animal success RATE, at full range
(0-9 both) and attempt-matched (current 0-3 vs previous full).
Finding: current > previous for both arms at full range (B2 0.54 vs 0.41
p=0.024; A2+Naive 0.42 vs 0.33 p=0.014) but the gap vanishes/reverses at
matched effort (B2 0.33 vs 0.41 n.s.; A2 0.30 vs 0.33 n.s.) -- a practice/
attempts artifact (~2.7x more attempts/session), not a stronger cohort.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add make_matched_effort.m: pick the current-study day cutoff whose per-animal
cumulative attempts best match the previous (_f) study's full-span total
(~405 attempts/animal -> current days 0-3), then fit the paper LME on both.
Two variation folders (data.csv + analyze.m + result.txt):
variations/prev_f_full/ b2_f vs a2_f, days 0-9 (24 rats)
variations/matched_current_d0_3/ Box-B2 vs Box-A2, 0-3 (6 rats)
At matched cumulative effort both show a significant positive stim x day
interaction (prev p=0.008 +0.56/day; current-0-3 p=0.004 +10.2/day) -- the
tDCS acceleration replicates at equal practice, though the count slopes are
not directly comparable across datasets given differing attempts/session.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add forouzan_load_data.m (joins raw/Data_Forouzan.csv + raw/Total_Success.csv)
and make_forouzan_data.m, producing analysis/forouzan_reach_data.csv in our
curated column format (subject, group, day, success, total). Group is mapped
Anodal -> b2_f (tDCS), Control -> a2_f (control); Hand ignored; day 0-indexed
(Session 1 = day 0 = paper 'Day 1'); early-stopped sessions (total=0) dropped.
Conversion validated by refitting the paper's own LME on the converted data:
interaction t(227)=2.66, F(1)=7.09, p=0.0083 -- reproduces the paper's
reported t(227)=2.68, F(1)=7.12, p=0.008, N=24 (12 Anodal / 12 Control).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add make_variations.m + variation_analyze.m generating 20 self-contained
subfolders under analysis/matlab/variations/, one per grouping x day-window,
each with exactly three files: data.csv (curated subset), analyze.m (a simple
standalone script fitting behavior ~ stim + day + stim:day + (1|rat) on the
success COUNT), and result.txt (its output). Plus variations/SUMMARY.csv.
Groupings (stim=1 / stim=0, other groups dropped):
unmerge B2 vs A2
right_only B2+Right vs A2 (Box-A dropped)
naive_a2 B2 vs A2+Naive (Right, Box-A dropped)
naive_boxa B2 vs A2+Naive+Box-A (Right dropped)
Windows: 0-10, 0-13, 0-5, 6-10, 6-13. All 20 have equal day coverage.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Make tdcs_paper_lme window-aware (windowKey 'full'|'d0_10'|'d0_13',
default 'full' -> unchanged filename). Windowed runs give the two anchor
arms equal day coverage, so the stim:day interaction is not confounded by
the full-range coverage imbalance; the report now states whether coverage
is equal and adds an INTERPRETATION block. Add switch cases
paper_mergeNaive_d0_10 / _d0_13 and include them in run_all.
Finding: unlike mergeA2 (fair-window interaction n.s.), the pooled-naive
control keeps the interaction significant in the fair d0_13 window
(F(1)=7.07, p=0.0088, +1.80 [+0.46,+3.14]; equal on Day 1 p=0.20),
reproducing the paper without the coverage confound; d0_10 is borderline
(p=0.051).
Tests: tLme asserts equal coverage / no warning / one full summary for
both windows and a significant positive d0_13 interaction; bad windowKey
errors. Suite 41/41.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add tdcs_model_summary as the single place that renders a fitted model
verbatim (disp(model) with the Command-Window <strong> markup stripped),
and route every report through it so all scenarios emit the complete
model-fitting output:
- tdcs_report: adds the learning-rate GLMM's full summary (count/rate
already had theirs); localCleanDisp now delegates to the shared helper.
- tdcs_lme_report / tdcs_paper_lme: embed the full fitlme summary before
the curated effect table.
- tdcs_phase_lme: appends each phase's full fitlme summary.
Tests: tReport asserts 3 native summaries in a GLMM report; tLme asserts
the lme_*/paper_*/phase_* reports embed 1/1/3 summaries with markup
stripped. Suite 39/39.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Document Box-A2 as an a-priori sham/sensory control and the two-tier
analysis it implies: B2-vs-A2 as the primary sensation-controlled contrast,
and the pooled-control mergeNaive grouping as a confirmatory power-boost
licensed by A2/Naive exchangeability (with the TOST caveat that the
exchangeability is supported but not proven at n=3-4).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds cfg.mergeNaive and scenario/paper/phase/power support. This pooled grouping
is the closest replication of the paper's full pattern: paper LME interaction
F(1)=7.30, p=0.008 (paper 7.12/0.008), stim main n.s. (p=0.21 = equal on Day 1),
strong day effect; the overall tDCS advantage is significant at the cluster-honest
per-animal level (rate p=0.003) and well powered. Adds a grouping test. Suite 38/38.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds tdcs_paper_lme replicating the paper's exact formula/variable names (paper_*
switch cases): mergeA2 reproduces the interaction F(1)=7.09 vs 7.12, p=0.009 vs
0.008, with a coverage-confound warning. Rewrites tdcs_power_sim to fit that same
formula (interaction term canonicalized to day:stim). Adds a replication test.
Suite 37/37.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds make_figure (2-panel learning curves + per-phase slopes, colorblind-safe
Box-B2/Box-A2 palette) -> results/figure_learning.png; analysis/writeup.md
(figure + methods + results summary); and tdcs_export_all/run_export exporting
every data variation into an organized export/ tree (curated_all, scenarios/9,
anchors/3 with tDCS factor, phases/9 + MANIFEST). Adds tExport tests. Suite 36/36.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds tdcs_phase_lme (same LME refit within learning phases 0-5/6-10/6-13, with
per-phase slopes + interaction 95% CI) as phase_* switch cases, and
tdcs_power_sim (Monte-Carlo power for the early-phase interaction, scored by the
cluster-honest per-animal test and the LME test) + run_power. Findings: the
Box-B2 faster-acquisition signal is in the early phase; at n=3-5/group honest
power is 0.3-0.7 even at the observed effect (need ~8/group if effect is as
observed, ~20/group if half). Adds tPhasePower tests. Suite 35/35.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Box-A2 has no data past ~day 13, so the full-range days x tDCS interaction
extrapolates A2's slope and is confounded (spuriously significant, wrong sign).
Add the _d0_13 window (equal anchor coverage) and a report warning when the two
groups' day coverage is unequal. In the fair 0-13 window the interaction is n.s.
with a POSITIVE slope diff and the Day-1 tDCS effect is n.s. -- matching the
paper's 'equal on Day 1' direction (underpowered at n=8). Suite 32/32.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The model already evaluates the tDCS main effect at day 0, which is the paper's
'Day 1'; relabel the report/docstring to state the equivalence explicitly so the
'equal on Day 1' comparison is unambiguous. No model change (re-indexing day
would move the main-effect evaluation point off Day 1). Suite 31/31.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Adds 6 switch cases replicating the paper's linear mixed model
success ~ day * tDCS + (1|subject) on the Box-B2(tDCS=1) vs Box-A2(tDCS=0)
arm, reporting the days x tDCS interaction, days, and tDCS effects as t/F/p.
mergeA2_full reproduces the paper's interaction (F=7.09 vs 7.12, p=0.009 vs
0.008); interpretation is sign-aware and honestly notes our data diverges from
the paper (Box-B2 ahead on day 1, gap narrows -- not 'equal on day 1,
accumulates'). Adds tLme tests. Suite 31/31.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
fitglme coefTest only offers observation-level DF (df2~=178), which overstates
the group x day interaction significance for this few-subject design (p~=0 in
every scenario). Replace the reported learning-rate result with a cluster-honest
per-animal slope test (per-subject OLS slope of success vs day, Welch + MWU,
B2 vs A2) -- now correctly parallel in all 6 scenarios, matching the Python GEE.
The GLMM F-test is kept as a flagged reference only. Adds a regression test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Poisson GEE / binomial rate / learning-rate models, subject random-intercept
sensitivity, unknown-group classification and merge scenarios, plus METHODS.md
and learning-curve plots. Data reconstructed from the experiment DB and
verified against the exported matrix.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Experiments DB Dev
samsam2610