SMR Literature Positioning
Use this to place the contribution in the methods literature, not the substantive one. SMR
reviewers are methodologists who often know the closest prior estimator, the original derivation, and
the competing approach in a neighboring discipline. A missed precedent is the fastest path to a
reject.
Position against methods, not findings
The literature review of an SMR paper answers "what is the closest method, and how is yours
different?" — not "what is known about the substantive topic." Structure the review as a small map:
- Direct ancestors: the method(s) your contribution extends, corrects, or replaces. Cite the
primary source, not a textbook summary.
- Cross-discipline siblings: the same problem solved in statistics, econometrics, psychometrics,
machine learning, or network science. SMR readers expect you to know that a "new" sociology method
may already exist under another name elsewhere.
- Competing methods: the alternatives a reviewer will demand you beat in simulation. Name them
here so the comparison set in
smr-simulation-studies is not a surprise.
The precedent audit
Run this before drafting the review:
- Search the original statistical literature, not just sociology — many SEM, latent-variable,
multilevel, missing-data, and causal-inference tools originate in statistics/psychometrics.
- For each near-neighbor, write one line: what they do, what you do differently, why the difference
matters. If you cannot state the difference, you have not yet differentiated the contribution.
- Distinguish re-derivation from novelty: presenting a known result as new is a credibility
killer at a methods venue. If you rediscovered something, say so and add what is genuinely new.
Cross-field translation table
| Sociology framing |
Likely prior literature to check |
Risk if missed |
| Causal effect with controls |
statistics (potential outcomes, DAGs), econometrics |
"good/bad controls" already settled elsewhere |
| Latent classes / trajectories |
psychometrics, biostatistics (finite mixtures, GBTM) |
reinventing a named model |
| Measurement invariance |
psychometrics (MGCFA, alignment) |
overstating a known non-invariance result |
| Missing data |
statistics (MI, IPW, FIML), biostatistics |
ignoring the standard estimator |
| Network effects |
network science, spatial econometrics |
a known identification problem |
| Text-as-data |
NLP, computational linguistics, comp. social science |
a method already standard in CSS |
Sibling-journal misattribution guard
When citing "where methods like this are published," be precise:
- Sociological Methodology (ASA annual) ≠ Sociological Methods & Research (SAGE). Do not
conflate them in the cover letter or positioning.
- Psychological Methods (APA), Political Analysis, Structural Equation Modeling, and the
Journal of Educational and Behavioral Statistics are neighbors, not SMR. Attribute exemplar
papers to the correct venue (see
resources/exemplars/library.md).
What an SMR reviewer checks first
A methodologist refereeing the positioning section typically does three things in order: (1) scans the
reference list for the primary derivation of your direct ancestor — if only a handbook chapter appears,
credibility drops before page two; (2) asks whether the strongest competitor is named and carried into
the simulation, because a review that names it and a simulation that omits it reads as evasion; (3)
checks the cross-discipline column — a reviewer trained in psychometrics or biostatistics will recognize
a renamed finite-mixture or IPW variant instantly. Write the review so each of these three probes finds
its answer within one paragraph, and state explicitly which neighboring literature you searched and
found empty, so the referee does not assume you never looked.
Checklist
Anti-patterns
- Substantive review in a methods paper: paragraphs on the topic instead of on the method.
- Sociology-only search: missing the statistics/psychometrics origin of the method.
- Textbook citation for a primary result: citing a handbook instead of the derivation.
- Strawman map: omitting the strongest competitor so the simulation looks favorable.
- Venue confusion: attributing a Sociological Methodology or Psychological Methods paper to SMR.
Output format
[Direct ancestors] <method -> your difference>
[Cross-discipline siblings] <field : closest prior + difference>
[Competing methods to beat] <named alternatives for the simulation>
[Precedent risks] <any near-rediscovery to disclose>
[Next SMR skill] smr-derivation-and-properties
1---2name: smr-literature-positioning3description: Use when positioning a Sociological Methods & Research (SMR) manuscript against the methods literature across sociology, statistics, econometrics, psychometrics, and computational social science, and avoiding sibling-journal misattribution. Maps the contribution onto prior methods; does not derive or simulate.4---56# SMR Literature Positioning78Use this to place the contribution in the *methods* literature, not the substantive one. SMR9reviewers are methodologists who often know the closest prior estimator, the original derivation, and10the competing approach in a neighboring discipline. A missed precedent is the fastest path to a11reject.1213## Position against methods, not findings1415The literature review of an SMR paper answers "what is the closest *method*, and how is yours16different?" — not "what is known about the substantive topic." Structure the review as a small map:1718- **Direct ancestors**: the method(s) your contribution extends, corrects, or replaces. Cite the19 primary source, not a textbook summary.20- **Cross-discipline siblings**: the same problem solved in statistics, econometrics, psychometrics,21 machine learning, or network science. SMR readers expect you to know that a "new" sociology method22 may already exist under another name elsewhere.23- **Competing methods**: the alternatives a reviewer will demand you beat in simulation. Name them24 here so the comparison set in `smr-simulation-studies` is not a surprise.2526## The precedent audit2728Run this before drafting the review:2930- Search the **original** statistical literature, not just sociology — many SEM, latent-variable,31 multilevel, missing-data, and causal-inference tools originate in statistics/psychometrics.32- For each near-neighbor, write one line: *what they do, what you do differently, why the difference33 matters*. If you cannot state the difference, you have not yet differentiated the contribution.34- Distinguish **re-derivation** from **novelty**: presenting a known result as new is a credibility35 killer at a methods venue. If you rediscovered something, say so and add what is genuinely new.3637## Cross-field translation table3839| Sociology framing | Likely prior literature to check | Risk if missed |40|---|---|---|41| Causal effect with controls | statistics (potential outcomes, DAGs), econometrics | "good/bad controls" already settled elsewhere |42| Latent classes / trajectories | psychometrics, biostatistics (finite mixtures, GBTM) | reinventing a named model |43| Measurement invariance | psychometrics (MGCFA, alignment) | overstating a known non-invariance result |44| Missing data | statistics (MI, IPW, FIML), biostatistics | ignoring the standard estimator |45| Network effects | network science, spatial econometrics | a known identification problem |46| Text-as-data | NLP, computational linguistics, comp. social science | a method already standard in CSS |4748## Sibling-journal misattribution guard4950When citing "where methods like this are published," be precise:5152- *Sociological Methodology* (ASA annual) ≠ **Sociological Methods & Research** (SAGE). Do not53 conflate them in the cover letter or positioning.54- *Psychological Methods* (APA), *Political Analysis*, *Structural Equation Modeling*, and the55 *Journal of Educational and Behavioral Statistics* are neighbors, not SMR. Attribute exemplar56 papers to the correct venue (see `resources/exemplars/library.md`).5758## What an SMR reviewer checks first5960A methodologist refereeing the positioning section typically does three things in order: (1) scans the61reference list for the primary derivation of your direct ancestor — if only a handbook chapter appears,62credibility drops before page two; (2) asks whether the strongest competitor is named *and* carried into63the simulation, because a review that names it and a simulation that omits it reads as evasion; (3)64checks the cross-discipline column — a reviewer trained in psychometrics or biostatistics will recognize65a renamed finite-mixture or IPW variant instantly. Write the review so each of these three probes finds66its answer within one paragraph, and state explicitly which neighboring literature you searched and67found empty, so the referee does not assume you never looked.6869## Checklist7071- [ ] Direct method ancestors are cited from primary sources, not textbooks.72- [ ] Cross-discipline siblings (statistics/psychometrics/econometrics/CSS) were searched and cited.73- [ ] Each near-neighbor has a one-line "what they do / what you do / why it matters."74- [ ] The simulation comparison set is foreshadowed in the review.75- [ ] No known result is presented as novel; re-derivations are flagged.76- [ ] Sibling journals are attributed correctly; no SMR/Sociological Methodology conflation.7778## Anti-patterns7980- **Substantive review in a methods paper**: paragraphs on the topic instead of on the method.81- **Sociology-only search**: missing the statistics/psychometrics origin of the method.82- **Textbook citation for a primary result**: citing a handbook instead of the derivation.83- **Strawman map**: omitting the strongest competitor so the simulation looks favorable.84- **Venue confusion**: attributing a *Sociological Methodology* or *Psychological Methods* paper to SMR.8586## Output format8788```text89[Direct ancestors] <method -> your difference>90[Cross-discipline siblings] <field : closest prior + difference>91[Competing methods to beat] <named alternatives for the simulation>92[Precedent risks] <any near-rediscovery to disclose>93[Next SMR skill] smr-derivation-and-properties94```
brycewang-stanford