Repeating a psychology paper’s headline can lose the sample, procedure and limits that make its claim meaningful. Read for a claim-and-evidence record, then write a conclusion that stays within what the authors studied.
To read a psychology study, connect its research question to its sample, design, measurements, results and limitations. Begin with the abstract as a map, verify details in the Methods and Results, and separate the authors’ interpretation from your own restatement. The sheet below includes a worked reading of a real public paper, not an invented study or a statistics tutorial. Missing or unclear information belongs in the record as a question.
- Find the full paper and identify the specific study within it.
- Record the recruited sample separately from the analyzed sample.
- Name the measured outcome rather than a broad psychological label.
- Keep the result, interpretation and remaining uncertainty in separate fields.
The Research Methods guide supplies concept routes when a design term is unfamiliar. The purpose here is to organize actual evidence; the existing research-methods practice questions serve the separate task of practicing design distinctions.
Orient to the actual paper before taking notes
Check the title, authors, publication details and full-text availability. Then identify whether the paper contains one study, several studies, a review or a commentary. An abstract is useful for orientation, but it cannot stand in for the sample details, exclusions or procedures contained in the full report.
For this worked example, read Craig Hedge, Georgina Powell and Petroc Sumner’s The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences, published in Behavior Research Methods. The record distinguishes the 2017 online publication from the 2018 volume publication. The paper reports three studies of repeated cognitive-task measurements, making it useful for seeing how an attractive broad conclusion depends on specific measures.
The abstract starts the map; the evidence fields are filled from the full paper.
A practical route is: read the abstract and section headings; inspect the Methods; identify the result relevant to your question; read the Discussion for interpretation and limitations; then return to the abstract. This is an editorial reading route, not a universal sequence required by every discipline. If a reported measure remains vague, use the operational-definition worksheet to identify what procedure is missing.
Takeaway: Choose the paper and the specific study before extracting a finding.
Fill the claim-and-evidence reading sheet
Use one sheet per claim or study, rather than combining every result into a single summary. Each field includes a location so another reader can check your note. If you cannot locate the information, record “not located” and the section you inspected rather than completing the cell from an assumption.
| Field | Your note | Location in the paper |
|---|---|---|
| Research question | _____ | _____ |
| Specific study and sample | _____ | _____ |
| Recruitment and exclusions | _____ | _____ |
| Design and comparison | _____ | _____ |
| Manipulated or recorded variables | _____ | _____ |
| Operational measure and units | _____ | _____ |
| Result relevant to the question | _____ | _____ |
| Statistic and uncertainty, if understood | _____ | _____ |
| Authors’ interpretation | _____ | _____ |
| Limitations and unresolved questions | _____ | _____ |
| My bounded restatement | _____ | _____ |
Illustrative photo — generated editorial image, not a real classroom study, participant or result.
Add a section or table locator to every factual note in your own reading record.
OpenStax’s approaches-to-research section explains why design affects what a study can establish. Its analyzing-findings section distinguishes operational definitions, assignment and interpretation. Keep these concepts available through the Research Methods resources while filling the sheet.
Takeaway: A checkable reading note contains both a statement and its locator.
Work through the public reliability paper
The paper asks whether established cognitive tasks provide reliable measurements of individual differences across repeated sessions. This differs from asking whether a condition effect appears in the group. A consistent group contrast and a consistent measurement of differences between people are not interchangeable outcomes.
| Sheet field | Worked entry from the original paper | Locator and boundary |
|---|---|---|
| Question | How well do commonly used cognitive-task measures retain individual differences across sessions? | Introduction, “Present study”; scoped to the examined tasks and measures |
| Recruited samples | Study 1: 50 undergraduate students; Study 2: 62; Study 3: 42 | Methods, “Participants”; these are initial samples, not every analysis denominator |
| Design | Participants complete task batteries in repeated sessions; task order is kept the same within a participant across sessions and counterbalanced across participants | Methods, “Design and procedure”; describe the reported procedure rather than calling it a treatment trial |
| Measures | Task response times, errors and task-specific contrasts; test–retest reliability summarized using intraclass correlations | Methods, task descriptions and “Data analysis”; the measure must be named before reading its coefficient |
| Analysis sample example | Stroop analyses retain 47 participants in Study 1 and 56 in Study 2 after the reported exclusions | Methods, “Data analysis”; do not replace those numbers with the recruited sample sizes |
| Result and interpretation | Expected experimental effects are observed, while reliability of individual-difference measures varies; many contrast measures have lower reliability than their component measurements | Results, “Task performance” and reliability tables; no one coefficient represents every task |
| Reading limit | The evidence concerns these tasks, scoring procedures, samples and repeated-session arrangements | Discussion; it does not validate a homemade classroom score |
The recruited counts above—50, 62 and 42 students—and the two retained Stroop counts—47 and 56 participants—come from the original paper’s Methods, checked October 8, 2026. They illustrate why a reader should not use a single “sample size” for all rows of a multi-study article. No other demographic counts or effect sizes are needed for this reading task.
Illustrative photo — generated editorial image, not a real classroom study, participant or result.
The paper example requires a study and a task locator before selecting the denominator. At a glance: Which sample is reported? · Recruited sample · Exclusions by task · Retained analysis sample · Locate study and task.
The Stroop classroom planning sheet applies a related boundary to a different task: organizing a teaching demonstration without ranking people. It is not the procedure used in this reliability paper.
Takeaway: Preserve the study, task and analysis sample attached to each result.
Read results without inventing statistics
Start with the comparison and measured quantity, then read the statistic and its uncertainty. A table of intraclass correlations describes measurement agreement across sessions under the authors’ analysis. It is not a table of percentages of people who “have good attention,” and it is not a clinical threshold.
Use a small result note: “Measure ___; comparison ___; estimate ___; uncertainty ___; table ___.” If you do not understand a coefficient, leave its interpretation unresolved while copying its exact label and locator. Do not translate an unfamiliar number into a dramatic everyday claim.
For this example, the worked summary intentionally does not select a single reliability coefficient. Selecting one would require naming the task, response measure, study and relevant interval. The published tables provide those distinctions; the broad abstract range is not a replacement for them.
Common mistake: Combining the largest recruited sample with a result from another task, then describing it as the paper’s single definitive finding.
The headline rewrite activity helps check the strength of the verb in your result sentence. The variables worksheet helps keep the recorded measure distinct from the construct it is meant to represent.
Takeaway: Leave an unfamiliar statistic open rather than inventing its practical meaning.
Write a bounded restatement you can defend
A useful restatement includes the population, procedure and measured conclusion. Here is an original paraphrase of this reading: “In undergraduate samples completing the authors’ task batteries across repeated sessions, established experimental contrasts did not uniformly provide reliable measurements of individual differences; suitability depended on the task and score.”
That sentence is narrower than “cognitive tests do not work.” It also does not mean a weak individual-difference score erases an experimentally observed contrast. The important reading distinction is between the research purposes and the measures supporting them.
A strong restatement identifies what the evidence concerns and avoids a universal headline. At a glance: Build a bounded sentence · Sample · Procedure · Measured finding · Limit.
Try writing your own sentence from the completed sheet before rereading the paper. The explain-and-check routine helps record whether you omitted the sample, confused the measure or overstated the scope. It provides a study log rather than new evidence for the paper.
Takeaway: Make every broad word in the summary traceable to a specific part of the record.
When this reading sheet is insufficient
The sheet helps beginners organize a report; it cannot replace expertise in a statistical method, a full systematic review or an evaluation of a clinical intervention. A single study cannot answer every population or application question. Linked supplementary files and datasets require their own inspection; this walkthrough does not claim to have reanalyzed them.
If the full text is unavailable, limit your record to the material you can actually read and identify the missing sections. Do not write a full-method critique based on an abstract alone. For another concept to study after the reading, use the Practice guide rather than expanding the restatement beyond its evidence.
Takeaway: Treat the sheet as a record of your reading, with visible unresolved questions.
Method and sources
The worked entries were checked in the cached full original paper, including Methods, Data analysis, Results and the reliability tables, accessed October 8, 2026. OpenStax Psychology 2e, sections 2.2 and 2.3, supplies the introductory design and measurement definitions. The sheet, reading route and final paraphrase are original editorial work, and the photos are generated illustrations; no new experiment, statistical reanalysis or tested reading benefit is claimed.
Takeaway: Keep your evidence record available alongside your final summary.
