Situational Evidence: Strategies for Causal Reasoning From Observational Field Notes
Abstract
There is unexamined potential for developing and testing rival causal explanations in the type of data that participant observation is best suited to create: descriptions of in situ social interaction crafted from the participants’ perspectives. By intensively examining a single ethnography, we can see how multiple predictions can be derived from and tested with field notes, how numerous strategies are available for demonstrating the patterns of nonoccurrence which causal propositions imply, how qualitative data can be analyzed to negate researcher behavior as an alternative causal explanation, and how folk counterfactuals can add to the evidentiary strength of an ethnographic study. Explicating the potential of field notes for causal explanation may be of interest to methodologists who seek a common logic for guiding and evaluating quantitative and qualitative research, to ethnographic fieldworkers who aim at connecting micro- to macro-social processes, to researchers who use an analogous logic of explanation when working with other forms of qualitative data, and to comparative–analytic sociologists who wish to form concepts and develop theory in conformity with an understanding that social life consists of social interaction processes that may be captured most directly by ethnographic fieldwork.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In Stepin et al.'s method [122], the generated counterfactuals are ranked using a XOR-based distance to find the most relevant counterfactual pertinent to the given contrast class.
“Subsequently, the generated counterfactuals are ranked using a XOR-based distance to find the most relevant counterfactual pertinent to the given contrast class.”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
Stepin et al. retrieve counterfactuals from a rule matrix where each rule is encoded in terms of all possible feature values [122].
“Stepin et al. retrieve counterfactuals from a rule matrix where each rule is encoded in terms of all possible feature values [122].”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
Several model-specific frameworks generate counterfactuals to explain the output of decision trees [46], [80], [122], [155].
“Several model-specific frameworks generate counterfactuals to explain the output of decision trees [46], [80], [122], [155].”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
Stepin et al.'s method [122] is further extended to generating counterfactuals for fuzzy decision trees.
“This method is further extended to generating counterfactuals for fuzzy decision trees.”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
Single-sentence textual counterfactual explanations suggest feature-value based instructions or alternative actions for a possible output change.
“Such explanations suggest featurevalue based instructions [48], [122], [131] or alternative actions for a possible output change [84], [149].”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
The most commonly acceptable definition of a counterfactual in the observed RQ3-related studies states that a counterfactual explanation is a set of minimal feature modifications that makes the model change the prediction.
“Nevertheless, the most commonly acceptable definition of a contfactual in the observed RQ 3 -related studies states that a contfactual explanation is a set of minimal feature modifications that makes the model change the prediction [81], [105], [122].”
From A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021) · cites this paper as [122]
The citation lists haven’t been fetched yet
References 62 · Cited by 30
- Authors
- Katz, J.
- Affiliations
- University of California, Los Angeles (US)
- Corresponding author
- Katz, Jack
- ORCID
- Katz, Jack
- Venue
- Sociological Methods & Research
- Publisher
- SAGE Publications
- Published
- 2014-11-23
- Volume
- 44
- Issue
- 1
- Pages
- 108-144
- ISSN
- 0049-1241 (ISSN-L)
- 1552-8294
- OpenAlex
- W2120257858
- Cited by
- 30
- References
- 62
- Citations by year
- 2026: 2 · 2025: 1 · 2024: 2 · 2023: 4 · 2022: 1 · 2021: 4 · 2020: 3 · 2019: 4 · 2018: 2 · 2017: 6 · 2015: 1
- FWCI
- 7.60
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Sociological Methods & Research (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- eScholarship (California Digital Library) (submittedVersion)
- Topics
- Social Sciences › Sociology and Political Science › Qualitative Comparative Analysis Research
- Social Sciences › Education › Qualitative Research Methods and Applications
- Social Sciences › Sociology and Political Science › Qualitative Research Methods and Ethics
- Keywords
- causal explanation, field notes, participant observation, counterfactual reasoning, researcher behavior, qualitative evidence, theory development
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
