Use when executing and reporting the analysis/interpretation for an American Anthropologist (AA) manuscript so it survives expert anonymous review — disciplined ethnographic/qualitative inference, honest interpretation, and (for biological/archaeological work) sound quantitative analysis. Guides analysis and interpretation norms; it does not fabricate evidence or quotations.
NewMaintained
Automatic
◆Research pack
Last verified 8/15/2026
What it changes
Passed 19 of 22 test cases with the skill, and 19 of 22 without it.
Without skill86%
With skill86%
Checks it passed that the model alone missed
Requires replacing generalized assertions with specific quoted fieldnote evidence
Identifies linguistic anthropology as the primary four-field subfield mode for interactional code-switching analysis
2 fixed17 passed either way1 still failing2 broke
Pass-rate lift
0 pp
Tokens
+41%
Agent turns
0%
Response time
-19%
Model gemini-3.6-flashMeasured 8/15/2026v119 of 22 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 2 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 2
Prompt
Here is a draft section analyzing code-switching and stance-taking in urban youth radio broadcasts. I want an analysis review that breaks down our findings into a structured report block. Make sure to specify the core finding, the grounding transcript/media evidence, the primary analytical subfield mode, how rival readings are handled, whether contradictory instances are addressed, ethical protection and attribution checks, positionality factors, and the follow-up step.
What was checked, without and with the skill
Checked
Without skill
With skill
Identifies linguistic anthropology as the primary four-field subfield mode for interactional code-switching analysis
Missed
Passed
Evaluates whether alternative readings of speech stances are explicitly addressed
Passed
Passed
Requires accounting for contradictory transcript instances in the evidence
Passed
Passed
case-03 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
AA reviewers are expert and read across four fields, so the same results section may be read by an
ethnographer, an archaeologist, and a bioanthropologist at once. At AA, interpretation is analysis:
qualitative inference is held to a real standard, not waved through as "context." This skill covers
execution and reporting norms; design decisions live in amanthro-research-design.
When to trigger
Building the results/interpretation section from fieldnotes, transcripts, material, or measurements
A reviewer asked for more evidence, alternative readings, or robustness
Reconciling what interlocutors say with what they do, or pre-planned vs. emergent themes
Making the analysis transparent enough to be credible without exposing protected sources
Analysis & interpretation norms AA expects
Show the evidence behind the claim. Ground each interpretive claim in specific, quoted, or
described evidence (a fieldnote moment, an utterance, an artifact, a measurement) — not assertion.
Pursue the alternative reading. Name the strongest rival interpretation and show why the
evidence favors yours; report where it doesn't fit cleanly. AA values honesty about ambiguity.
Disconfirming evidence. Include cases that complicate the argument rather than curating only
confirming material; say what you learned from the friction.
Quote and represent fairly. Translate and contextualize interlocutors' words; do not flatten
them into illustrations of your theory. Attribute analytics to communities where they originate.
For quantitative subfields (biological/archaeological): report effect sizes and uncertainty, not
just significance; pre-specify where possible; correct for multiple comparisons; validate measures;
avoid conflating ancestry with social race.
Reflexivity in the analysis. Note how your position shaped which evidence was available and how
you read it.
Mixed-methods / computational specifics
If you code text or use computational tools on field/archival data, document the procedure, validate
against close reading, and treat outputs as interpretable evidence, not ground truth.
Keep the qualitative core legible: a table or model never replaces the interpretive argument at AA.
AA four-field evidence-and-care ledger
Before writing the results section, build an AA ledger that shows how the analysis earns trust across
anonymous four-field review while protecting the people, places, materials, and communities that make
the evidence possible.