Canadian Law and Society Association
ENFR
CLSA / ACDS

An interdisciplinary
conversation on law & society

research

Read an empirical article without losing its limits

A reusable reading record with a worked example from Crosby and colleagues’ 2025 study of police-involved deaths in Canada.

What can you conclude from an abstract?

An abstract can identify the research question, reported material and principal finding. It cannot replace checking the full methods or underlying evidence. In your reading record, label what you actually read, keep the authors’ finding separate from your interpretation, and leave unexamined fields unresolved.

Turn a sociolegal theme into a researchable question · Read a membership survey without losing the denominator

Make a record you can check again

Read once for the argument and again for the evidence. Keep the authors’ finding, your interpretation and your next question in separate fields. Add a page or section locator so a seminar participant can check the passage without guessing.

An abstract is an entry point. It does not establish that you have read the full methods, coding choices, uncertainty analysis or underlying data. Mark those fields as pending until you inspect them.

Five passes through the article

  1. Identify the question and the unit of analysis: people, cases, decisions, institutions, interviews or another defined unit.
  2. Record place, period, inclusion criteria and sources. Note what the material excludes or cannot observe.
  3. Describe the method in a sentence that another reader can understand. Separate collection from analysis.
  4. Locate the principal finding. Distinguish an observed pattern from a proposed explanation or a causal conclusion.
  5. Write one limitation and one follow-up source. Check the full article before presenting an abstract-level note as a full review.

Worked record: Tracking (In)Justice

Andrew Crosby and eight co-authors published “Tracking (In)Justice: Documenting Fatal Encounters with Police in Canada” in CJLS / RCDS 40(1), pages 23–47, in 2025. The record below paraphrases its abstract; it is not an independent replication.

Abstract-level reading record
QuestionHow can deaths involving police use of force across Canada be documented when existing knowledge is incomplete?
MaterialThe abstract reports 745 cases involving intentional police force in Canada, 2000–2023, assembled from publicly available sources.
ApproachA collaborative, interdisciplinary database and data governance that includes affected communities and families.
Reported patternThe authors describe rising deaths, especially shootings, and overrepresentation of Black and Indigenous people.
BoundaryThis historical case series is not a count of all police contacts, a present-day total or an individual risk forecast.
Next checkRead the article’s definitions, collection process and treatment of gaps before comparing rates or reproducing totals.

Copy this blank record into your notes

Reusable reading template
ReferenceAuthor(s), exact title, year, journal or publisher, DOI or stable URL.
QuestionWhat does the article ask?
MaterialUnit, place, period, sources, inclusions and exclusions.
MethodCollection and analysis; what is directly observed?
Claim + locatorFinding in your own words; page or section.
LimitWhat does the design leave unresolved?
My next stepOne question and the source needed to answer it.
A claim must remain connected to its research question, the material collected and the method’s uncertainty.
Use the steps alongside the sources below.

A claim must remain connected to its research question, the material collected and the method’s uncertainty.

Download a blank reading record

Download the editable UTF-8 text worksheet

The file contains the eight fields above, including whether you read an abstract or the full article. Open it in a text editor, enter your notes and save it locally. It contains no extracted journal text, no filled-in research result and no form submission.

Run a 50-minute article discussion with an explicit access limit

A seminar needs a shared evidence boundary before it needs a lively debate. This proposed 50-minute format is an adaptable teaching plan, not an ACDS–CLSA rule or a report of a session already held. It works with the article-reading record above and keeps unresolved methodological questions visible.

Prepare the material that everyone can inspect

For a concrete bibliography example, Crossref’s publisher-deposited record identifies Tracking (In)Justice: Documenting Fatal Encounters with Police in Canada, by Andrew Crosby and co-authors, Canadian Journal of Law and Society 40(1), 2025, pp. 23–47, DOI 10.1017/cls.2025.1. The accessible metadata includes an abstract. This check did not obtain the full publisher PDF. Do not treat the abstract as if it contained every inclusion rule, limitation or table from the complete article.

  1. Minutes 0–5: establish access. The facilitator records the citation and asks which participants have the full text. List the shared material precisely: abstract, bibliographic metadata and any legitimately available full article. Nobody needs to disclose personal information.
  2. Minutes 5–15: state one claim. A reader selects one sentence from the material actually available and supplies a locator. Another reader paraphrases it. Record whether the locator is in the abstract or a numbered full-text page.
  3. Minutes 15–25: map the evidence. Identify what the accessible passage says about scope and material. Put methodological details absent from it into a “requires full text” column. Do not fill those cells from the title.
  4. Minutes 25–35: formulate a challenge. Ask one question that could change the interpretation: for example, how cases were included or excluded. State the document or passage needed to answer it. A vote cannot establish the missing method.
  5. Minutes 35–45: compare readings. Separate disagreement about wording from disagreement about evidence. Keep two interpretations if the shared text cannot distinguish them; record their different assumptions.
  6. Minutes 45–50: close the record. Assign a next reading task, the required source and a review point. Check that the agenda totals 50 minutes: 5 + 10 + 10 + 10 + 10 + 5.

Completed model outcome: citation checked; shared source = abstract and metadata; selected claim locator = abstract; detailed inclusion/exclusion rules = not verified; decision = defer methodological evaluation; next task = obtain and read the relevant full-text methods; review = next discussion. This is a model record, not a finding about the article’s quality.

Save the blank version as: citation ___; shared access ___; claim/locator ___; interpretation ___; challenge ___; missing passage ___; next reader/task ___; review date ___. End with a traceable question, not an invented consensus.

The 50-minute agenda has six stages. An abstract-level claim remains separate from a question requiring full methods.
A proposed seminar plan, not a session result. Full publisher PDF was not obtained in this source check. © ACDS–CLSA · CC BY 4.0.

Crossref · publisher-deposited metadata and abstract · DOI 10.1017/cls.2025.1