Data Analysis (aaag-data-analysis)
The Annals expects analyses that are spatially honest and reported with uncertainty, whatever the
area. The standard is that a competent reader in the area could follow the logic from data to claim and
see that the geography of the data was respected, not flattened.
When to trigger
- Estimating models, running spatial statistics, classifying imagery, or coding qualitative material
- A reviewer questioned uncertainty, robustness, spatial autocorrelation, accuracy, or interpretation
- Preparing the results section and deciding what to report
Spatial / quantitative
- Diagnose space first. Report spatial autocorrelation in residuals; if present, move to a spatial
model (lag/error, GWR/MGWR, spatial regimes) and say why.
- Uncertainty everywhere. CIs/SEs (spatially robust where needed), not stars alone; for prediction,
out-of-sample error from spatial/blocked CV.
- Robustness. Re-estimate across plausible areal units and bandwidths (MAUP/scale sensitivity);
show the result is not a unit artifact. Report effect sizes in interpretable units.
Remote sensing / physical
- Accuracy with an independent sample. Confusion matrix, overall/producer/user accuracy, kappa or
F1; for continuous outputs, RMSE/MAE and bias; map the spatial pattern of error, not just a scalar.
- Propagate uncertainty from inputs through to the reported quantity; state the validation design.
Qualitative / interpretive
- Transparent analytic trail. Coding scheme, how themes were derived, and how interpretations were
checked (negative cases, member checks, triangulation) — credibility over counting.
- Evidence-to-claim mapping. Each interpretive claim is tied to identifiable (anonymized) evidence;
avoid quote-mining that over-generalizes from one informant.
Mixed methods
- Show the integration. State where the strands converge and where they conflict, and how the
conflict was adjudicated — do not report two parallel analyses and call it mixed methods.
Cross-cutting reporting bar
- Match every claim in the text to an exhibit or statistic; no orphan assertions.
- Report negative / null / scale-dependent results honestly; geography rewards scope conditions.
- Keep analysis reproducible: master script, seeds, pinned versions (see
aaag-transparency-and-data).
Referee pushback → Annals-specific fix
- "Are these effects just spatial autocorrelation?" → Show residual Moran's I before/after a spatial
model; report the spatial-error structure, not only a global coefficient.
- "Would the result change at a different scale/unit?" → Provide a MAUP/bandwidth sensitivity panel
and state the scale at which the claim holds.
- "How accurate is the map?" → Area-adjusted accuracy from an independent sample + a map of where
error concentrates, not a single kappa.
- "How do I know the qualitative reading isn't cherry-picked?" → Coding scheme, negative cases, and an
excerpt-to-claim table.
Calibration anchors
- Uncertainty is mandatory, not optional. A coefficient or accuracy number without an interval is
not yet a finding at this venue.
- Scale dependence is a result, not a nuisance. If the answer changes with the unit, say so — that
is geographic knowledge.
- The spatial pattern of error is itself a finding for remote-sensing and prediction work.
Checklist
Anti-patterns
- Reporting OLS on spatial data with no autocorrelation check
- Stars-only tables with no effect sizes or CIs
- A single global accuracy number with no spatial error map
- Cherry-picked quotes standing in for an analytic trail
- "Mixed methods" that never integrate the strands
Output format
【Mode】spatial-quant / remote-sensing / qualitative / mixed
【Headline result】effect/accuracy/theme + its uncertainty
【Spatial honesty】autocorrelation / MAUP / spatial-CV / spatial error map handled? [Y/N]
【Robustness】checks run and what held
【Reproducibility】master script + seeds + versions? [Y/N]
【Next】aaag-tables-figures
Supplementary resources
1---2name: aaag-data-analysis3description: Use when running and reporting the analysis for an Annals of the American Association of Geographers manuscript — spatial statistics and modeling, remote-sensing accuracy, or qualitative coding and interpretation. Sets analysis and reporting norms across the four areas; it does not choose the design.4metadata:5 source: https://skillmd.com/skills/brycewang-stanford/aaag-data-analysis6---78# Data Analysis (aaag-data-analysis)910The Annals expects analyses that are **spatially honest** and reported with uncertainty, whatever the11area. The standard is that a competent reader in the area could follow the logic from data to claim and12see that the geography of the data was respected, not flattened.1314## When to trigger1516- Estimating models, running spatial statistics, classifying imagery, or coding qualitative material17- A reviewer questioned uncertainty, robustness, spatial autocorrelation, accuracy, or interpretation18- Preparing the results section and deciding what to report1920## Spatial / quantitative21- **Diagnose space first.** Report spatial autocorrelation in residuals; if present, move to a spatial22 model (lag/error, GWR/MGWR, spatial regimes) and say why.23- **Uncertainty everywhere.** CIs/SEs (spatially robust where needed), not stars alone; for prediction,24 out-of-sample error from **spatial/blocked CV**.25- **Robustness.** Re-estimate across plausible **areal units and bandwidths** (MAUP/scale sensitivity);26 show the result is not a unit artifact. Report effect sizes in interpretable units.2728## Remote sensing / physical29- **Accuracy with an independent sample.** Confusion matrix, overall/producer/user accuracy, kappa or30 F1; for continuous outputs, RMSE/MAE and bias; map the **spatial pattern of error**, not just a scalar.31- **Propagate uncertainty** from inputs through to the reported quantity; state the validation design.3233## Qualitative / interpretive34- **Transparent analytic trail.** Coding scheme, how themes were derived, and how interpretations were35 checked (negative cases, member checks, triangulation) — credibility over counting.36- **Evidence-to-claim mapping.** Each interpretive claim is tied to identifiable (anonymized) evidence;37 avoid quote-mining that over-generalizes from one informant.3839## Mixed methods40- **Show the integration.** State where the strands converge and where they conflict, and how the41 conflict was adjudicated — do not report two parallel analyses and call it mixed methods.4243## Cross-cutting reporting bar4445- Match every claim in the text to an exhibit or statistic; no orphan assertions.46- Report **negative / null / scale-dependent** results honestly; geography rewards scope conditions.47- Keep analysis reproducible: master script, seeds, pinned versions (see `aaag-transparency-and-data`).4849## Referee pushback → Annals-specific fix5051- *"Are these effects just spatial autocorrelation?"* → Show residual Moran's I before/after a spatial52 model; report the spatial-error structure, not only a global coefficient.53- *"Would the result change at a different scale/unit?"* → Provide a MAUP/bandwidth sensitivity panel54 and state the scale at which the claim holds.55- *"How accurate is the map?"* → Area-adjusted accuracy from an independent sample + a map of where56 error concentrates, not a single kappa.57- *"How do I know the qualitative reading isn't cherry-picked?"* → Coding scheme, negative cases, and an58 excerpt-to-claim table.5960## Calibration anchors6162- **Uncertainty is mandatory, not optional.** A coefficient or accuracy number without an interval is63 not yet a finding at this venue.64- **Scale dependence is a result, not a nuisance.** If the answer changes with the unit, say so — that65 is geographic knowledge.66- **The spatial pattern of error is itself a finding** for remote-sensing and prediction work.6768## Checklist6970- [ ] Spatial autocorrelation diagnosed and addressed (quant)71- [ ] Uncertainty reported (CIs/SEs; out-of-sample error via spatial CV where relevant)72- [ ] MAUP/scale or bandwidth sensitivity shown (quant)73- [ ] Accuracy via independent validation + spatial error map (RS)74- [ ] Coding scheme + evidence-to-claim trail (qual); integration shown (mixed)75- [ ] Every textual claim maps to an exhibit/statistic7677## Anti-patterns7879- Reporting OLS on spatial data with no autocorrelation check80- Stars-only tables with no effect sizes or CIs81- A single global accuracy number with no spatial error map82- Cherry-picked quotes standing in for an analytic trail83- "Mixed methods" that never integrate the strands8485## Output format8687```88【Mode】spatial-quant / remote-sensing / qualitative / mixed89【Headline result】effect/accuracy/theme + its uncertainty90【Spatial honesty】autocorrelation / MAUP / spatial-CV / spatial error map handled? [Y/N]91【Robustness】checks run and what held92【Reproducibility】master script + seeds + versions? [Y/N]93【Next】aaag-tables-figures94```9596## Supplementary resources9798- [`../../resources/external_tools.md`](../../resources/external_tools.md) — spatial-stats, RS, and qualitative-analysis packages99- [`../../resources/README.md`](../../resources/README.md) — shared reporting-standards background (inference, robustness)