Introduction: What "Empirical" Means — and Why the Genre Has Rules

An empirical research paper reports original investigation: it answers a question with data the authors collected or analyzed — surveys, experiments, interviews, archival records, sensor readings — rather than with arguments assembled from other people's published conclusions. That single feature gives the genre its rigid, instantly recognizable structure (introduction, methods, results, discussion) and its unforgiving standards: every claim must trace to data, every method must be replicable, and every limitation must be acknowledged.

Students meet this genre in upper-division courses, theses, and first submissions to undergraduate research journals. It rewards writers who understand that the structure is not bureaucratic ceremony — it is the argument. The IMRaD frame mirrors the logic of science itself: here is what we did not know, here is what we did, here is what we found, here is what it means. Master the frame and the paper almost writes its own outline. This guide walks through the process step by step, from question to submission-ready draft.

Step 1: Move from Topic to Researchable Question

The Question Audit

An empirical question must be answerable with observable evidence. "Is social media bad?" is not empirical — "bad" is unmeasured. "Does daily TikTok use exceed 90 minutes correlate with lower self-reported academic focus among undergraduates?" is empirical: both variables are measurable and the relationship is testable. Audit your question for:

1. Variables you can actually measure with instruments you can actually access.
2. A population you can actually sample within your course or semester timeline.
3. An ethics path — human subjects require institutional review (IRB) or course-level ethical approval before any data collection begins.

Hypotheses Before Data

State your hypothesis (or hypotheses) in writing before you look at any data: a specific, falsifiable prediction, e.g., "Students who use a spaced-repetition app will score higher on a retention test than students who cram." Pre-registering your expectations — even informally, in a lab notebook or dated document — is the backbone of credible research.

Step 2: Design the Study and Write the Method Section First

Choose the Design That Fits the Question

• Experimental: you manipulate a variable (assign some students to the app, others to cramming) — the strongest design for causal claims.
• Quasi-experimental or correlational: you measure variables without manipulating them — appropriate when manipulation is impossible or unethical, but claims must be phrased as association, not causation.
• Qualitative (interviews, focus groups, observations): you collect language and behavior rather than numbers — analysis is thematic, and results are quoted rather than tabulated.

Write the method section before collecting data. If you cannot describe your procedure precisely enough for a peer to replicate it, you are not ready to collect anything.

The Method Section Checklist

• Participants: who, how many, how recruited, how sampled, and any exclusions.
• Materials/instruments: surveys, tests, or equipment — including where published instruments came from and their reliability if known.
• Procedure: the exact sequence, timing, and instructions participants experienced.
• Analysis plan: which statistical tests (or coding approach) you will use, and why.

Step 3: Collect Data Ethically and Keep Everything

Once your design and consent procedures are approved, collect data systematically. Practical rules that save papers later:

• Use consistent, pre-labeled storage for every file, with a codebook explaining every variable and value.
• Record everything you did that deviated from the plan — a missed session, a broken link, a participant who skipped questions. Deviations belong in your method section or limitations, not in your memory.
• Back up raw data in two places. Your analysis should always start from an untouched copy of the raw file.

Step 4: Analyze — and Let the Results Section Report, Not Argue

Run the Analysis You Planned

Analyze according to your stated plan. If you explore beyond it (common and legitimate), label those analyses as exploratory in the paper. The integrity of an empirical paper rests on the distance between what was predicted and what was merely noticed.

The Results Section's Golden Rule

Results sections describe, they do not explain. Report:

1. Descriptive statistics first — means, standard deviations, counts, response rates — in text, tables, or figures.
2. Inferential results second — the test used, the test statistic, degrees of freedom, the p-value, and the effect size, in the reporting format of your style guide (APA, for example, has exact templates).
3. Figures that carry weight: every table and graph must be referred to in the text, numbered in order, and interpretable without reading the prose.

Notice what is absent: no interpretation, no comparison to other studies, no "this suggests." Those sentences belong to the discussion. Writers who blur this line end up reorganizing half the paper at revision.

For qualitative work, the parallel rule holds: results present themes with supporting quotes; discussion interprets the themes.

Step 5: Write the Discussion — Where the Paper Earns Its Grade

The Four-Move Discussion

A strong discussion performs four moves in order:

1. Restate the finding in plain language — the answer to your research question, in one paragraph, with effect sizes or themes in plain view.
2. Connect to the literature — how the result aligns with, extends, or contradicts the studies in your introduction. A contradiction is not a failure; it is a finding worth a paragraph of explanation.
3. Confront limitations honestly — sample size, sampling bias, measurement limits, confounds you could not control. Rank them by how much they threaten your conclusion.
4. State implications and future directions — what a practitioner, researcher, or policymaker should do differently because your result exists.

The Introduction (Yes, Now You Write It)

With the analysis complete, the introduction writes itself to spec:

• Paragraph 1: the territory — the broad problem and why it matters.
• Paragraph 2: the gap — what previous research established and left unresolved (this is your literature review, compressed to the studies that set up your question).
• Paragraph 3: the present study — your research question, hypothesis, and a one-sentence preview of the approach.

Write the title last: it should name the variables and the population, e.g., "Spaced Repetition and Retention: A Comparison of App-Based and Cramming Study Schedules in Undergraduates."

Step 6: Format, Cite, and Revise Like a Researcher

• Cite as you write, in your assigned style (APA dominates the empirical sciences), with a reference manager keeping the library clean.
• Report statistics to convention: exact p-values (p = .03) unless below reporting thresholds, effect sizes for every inferential test, and no "trend toward significance" language for results that missed the threshold.
• Revise in passes: one pass for argument structure, one for data-reporting accuracy (check every number against your output files), one for prose.
• Read the method section aloud: awkward phrasing hides procedural ambiguity, and ambiguity in methods is what reviewers circle.

The Most Common Student Mistakes

• Claiming causation from correlation. If you did not manipulate and randomly assign, the word "cause" may not appear.
• Hiding inconvenient results. Reporting that one hypothesis was not supported increases credibility — and is required by research ethics.
• Method ambiguity. "Participants completed a survey" describes nothing. Which survey, how long, what scale, what instructions?
• Discussion that introduces new results. If a finding appears only in the discussion, it is misplaced — move it to results.
• Overlong literature reviews. Twenty background citations that never converge on your gap read as padding. Every source must do work that sets up your question.

Conclusion: The Structure Is the Argument

The empirical research paper is a genre where form enforces honesty: the introduction justifies the question, the method makes it repeatable, the results constrain what you may claim, and the discussion pays the reader back with meaning. Move through the sequence deliberately — question, hypothesis, design, methods-first writing, ethical collection, planned analysis, reporting results without interpretation, then a four-move discussion — and the genre's rigidity stops being a burden and becomes your scaffolding.

Your next step: write your research question and hypothesis on one line, then draft your method section's participants and procedure paragraphs before you collect a single data point. Every hour spent there saves a week of revisions.