ATLASResearch
methods

Atlas training · For students and academics

Guided training

Work through one lesson at a time. Try the tools, practise with an example and write a research decision brief you can explain.

8 lessons · About 35 minutes, plus practice

Prefer a printable booklet?

Prefer a booklet? Save all lessons, examples and your brief as a PDF using your browser’s print dialog.

Lesson 1 of 8 · About 4 minutes

Start with a research question

By the end: Turn a broad topic into a question that can guide a method choice.

Atlas is a teaching and reference tool for students, supervisors, lecturers and researchers. Use it when planning a proposal, preparing a supervision meeting, learning an unfamiliar method or comparing approaches. It connects philosophies, designs and techniques across six layers. It does not collect or analyse your data, write a dissertation, approve ethics or certify a design.

  1. 1

    Name the question, people and setting

    Write one sentence beginning with how, what, why, whether or to what extent. Specify the people, organisation, material or system you intend to study. Start with the question rather than a favourite method or software package.

    What you should have: A bounded question, such as: How do first-year commuter students describe belonging during their first semester?

  2. 2

    Decide what an answer would look like

    Would an answer explain experiences, describe a pattern, estimate an effect, predict an outcome, develop something or bring existing studies together? Different aims may require different evidence. Separate a question about association from a claim about cause.

    What you should have: A plain-language purpose: I want to understand students’ accounts of belonging.

  3. 3

    List the access you actually have

    Record the evidence you could obtain, permissions still needed, your skills and the time available. Access to a class does not automatically mean permission to use its records or recruit its students.

    What you should have: An honest starting constraint: potential interview access, subject to institutional approval and consent.

A concrete example

“Student wellbeing” is a topic. “How do postgraduate students describe managing assessment pressure during a taught semester?” is a question. “Does a support programme reduce stress?” asks something different and needs a design capable of supporting that claim.

Pause and explainYou want to know why students avoid a service, but all you currently have is an attendance total. What is missing?Reveal reasoning

Attendance totals can describe use. They do not by themselves explain students’ reasons. Revise the question to fit the evidence, or plan an appropriate way to investigate experiences and explanations, with permission.

Watch for this: Do not treat a method that appears first in Atlas as the correct answer. The output you need is a reasoned shortlist and a plan for checking it.

Open Find a method

The practice link opens in a new tab. Return to this tab to keep your place in the lesson.

Step 2 · Practise what you learned

Try a worked example.

These are fictional teaching scenarios, not reported study results or ready-to-run research protocols. Allow about 15 minutes for one. Entry links open in a new tab so you can keep this guide open. Open the named entries, then explain your reasoning before revealing the discussion point.

QUALITATIVE / TEACHING SCENARIO

Belonging at university

How do first-year commuter students describe belonging during their first semester?

A student has potential access to volunteers at one university, subject to approval. The interest is in experiences and meanings, not estimating how common each experience is across all universities.

Finder practice inputs

  1. PurposeUnderstand meaning and experience
  2. EvidenceText & talk
  3. DisciplineEducation
  4. TimeMedium

Inputs are for this scenario. Use your real constraints for your own project.

  1. Use Find a method with the teaching inputs shown here. Read the reasons in the current results rather than expecting a fixed winning method.
  2. Open Phenomenology and Qualitative case study in Method library. Add both to Compare and ask whether the focus is lived experience or a bounded institutional case.
  3. Open Semi-structured interviews separately as a possible collection technique. Read In practice, including access and assumptions. A technique does not decide the design for you.
  4. Read a source and complete a decision brief with your chosen focus, a feasibility issue and a limit on the claims.
PhenomenologyQualitative case studySemi-structured interviews

One line of reasoning to investigate

If the central concern is a bounded support programme in its university context, a case study may be worth investigating. Define the case and its boundaries, and consider which evidence sources would illuminate it. If the concern is the meaning of a lived experience, investigate the relevant phenomenological tradition and its assumptions.

Read the Qualitative case study entry and its Sources tab
Change one thingThe student changes the question to “What proportion of commuter students feel they belong?” What should change?Reveal discussion

The purpose now involves describing a distribution. Reconsider the sampling, measures and evidence needed for a descriptive study. A few qualitative interviews cannot establish the proportion for a whole population.

QUANTITATIVE / TEACHING SCENARIO

Study time and assessment

What is the relationship between weekly study time and assessment scores in a defined student cohort?

A researcher is considering numerical questionnaire measures and permitted assessment records. The exercise concerns an observational relationship, not a demonstrated effect of extra study time.

Finder practice inputs

  1. PurposeExplain a process or relationship
  2. EvidenceNumbers
  3. DisciplineEducation
  4. TimeMedium

Inputs are for this scenario. Use your real constraints for your own project.

  1. Use Method library to locate Cross-sectional survey and inspect its Overview and In practice tabs. Identify the population, timing and proposed measures.
  2. Open OLS and multiple linear regression. Ask whether the outcome, data structure and model assumptions suit your question. Read its assumptions before proposing it.
  3. Keep design and analysis decisions separate in the worksheet. A regression model is not a substitute for planning sampling, access and measurement.
  4. Open the published sources and write a cautious conclusion template. Note plausible alternative explanations and avoid inventing results.
Cross-sectional surveyOLS and multiple linear regressionRandomised experiment

One line of reasoning to investigate

A survey design and a regression analysis may contribute to the same project, if justified. A statistical association alone does not show that study time causes a score change. Students may differ in prior knowledge, workload and other relevant characteristics; your proposed design and analysis need to address the claims you intend to make.

Read the Cross-sectional survey entry and its Sources tab
Change one thingA significant association is found. Can the researcher say that increasing study time will improve every student’s score?Reveal discussion

No. That would exceed an observational association and also make a universal individual claim. Reconsider the causal question and a suitable design, such as the designs discussed in Atlas’s causal entries, before making an effect claim.

MIXED METHODS / TEACHING SCENARIO

Explaining feedback experiences

What patterns appear in students’ ratings of feedback, and how do students explain the unexpected patterns?

An academic team could gather numerical ratings and interview accounts, subject to approval. They need to plan recruitment across phases and explain how the two strands inform each other.

Finder practice inputs

  1. PurposeExplain a process or relationship
  2. EvidenceText & talk + Numbers
  3. DisciplineEducation
  4. TimeExtended

Inputs are for this scenario. Use your real constraints for your own project.

  1. Open Find a method and select both evidence types. Mixed methods requires access to both strands and a reason for combining them.
  2. Compare Explanatory sequential and Convergent. Read the order and relationship of the strands in each design, then the planning and access requirements.
  3. For an explanatory sequence, state how an unexpected numerical pattern would inform the next phase’s sampling and interview questions. Do not choose follow-up participants before you know what needs explaining.
  4. Read the integration source linked in the profile. Write down where the strands connect and how you will handle accounts that complicate the numerical pattern.
Explanatory sequentialConvergentSemi-structured interviews

One line of reasoning to investigate

An explanatory sequential approach may fit a numerical phase followed by qualitative work designed to explain it. A convergent approach involves a different relationship between the strands. Collecting a survey and interviews is not enough by itself: the project needs an explicit integration plan.

Read the Explanatory sequential entry and its Sources tab
Change one thingThe survey and interviews seem to tell different stories. Should the team remove the inconvenient strand?Reveal discussion

No. Examine sampling, timing, measures and context, and consider what the difference reveals. Report and interpret the disagreement transparently rather than forcing the strands into apparent agreement.

FINANCE / TEACHING SCENARIO

Market response to an announcement

How did a defined set of listed firms’ returns differ from an expected-return benchmark around an announcement?

A finance student has potential access to adjusted daily prices, a market series and dated announcements. The exercise concerns a short event window and a justified benchmark, not a claim that any price change must have been caused by the announcement.

Finder practice inputs

  1. PurposeExplain a process or relationship
  2. EvidenceTime series
  3. DisciplineBanking & finance
  4. TimeMedium

Inputs are for this scenario. Use your real constraints for your own project.

  1. Try the Finder inputs, then use Method library to inspect Event study even if it is absent from the six-item shortlist. The shortlist is selective, not an exhaustive search.
  2. Read the Event study prerequisites. Specify the event date, an estimation window, an event window and the return benchmark before inspecting the event result.
  3. Open CAPM as one benchmark model to investigate, and CRSP and WRDS as a data-access entry. These complement an event study rather than replace the design.
  4. Check data adjustments, overlapping events, other news and inference assumptions. Record a claim boundary and a source still to read, without inventing returns or test results.
Event studyCAPMFinancial research databases: CRSP and WRDS

One line of reasoning to investigate

An event study may help investigate abnormal returns around a defined event. A benchmark supplies the expected-return comparison; it is not proof of the counterfactual. A time-window association needs careful interpretation, particularly when other events overlap.

Read the Event study entry and its Sources tab
Change one thingA major earnings release occurred in the same event window. Can the whole abnormal return be attributed to the announcement of interest?Reveal discussion

No. The competing news complicates attribution. Investigate event contamination and defensible design changes, and narrow the claim if the evidence cannot distinguish the announcements.

COMPUTING / TEACHING SCENARIO

Can students use the new submission interface?

Where do students encounter difficulties when completing a coursework submission in a prototype interface?

A computing project has a prototype and potential access to consenting volunteers, subject to approval. It needs observed interaction and participant accounts, rather than a sales forecast or a measure of learning gains. Use fictional submissions, not real coursework or credentials.

Finder practice inputs

  1. PurposeEvaluate a programme, system or practice
  2. EvidenceText & talk + Numbers
  3. DisciplineComputer science
  4. TimeShort

Inputs are for this scenario. Use your real constraints for your own project.

  1. Try the Finder inputs. Inspect Usability testing, then describe realistic tasks, intended users and the prototype version.
  2. Separate observed difficulties and task outcomes from participants’ explanations. Decide how facilitation, recording and participant welfare will be handled.
  3. Compare Usability testing with Heuristic evaluation. The former observes participants attempting tasks; the latter uses expert inspection. Explain whether one or both address your evaluation question.
  4. Inspect System Usability Scale as a possible supplementary measure. A questionnaire score does not replace observed task evidence. Read the sources and record an achievable next evaluation step.
Usability testingHeuristic evaluationSystem Usability Scale

One line of reasoning to investigate

A task-based usability evaluation can reveal interaction problems within a defined prototype and participant group. Expert inspection can contribute a complementary perspective. Neither by itself establishes that the interface improves academic attainment or represents every future user.

Read the Usability testing entry and its Sources tab
Change one thingThe prototype receives a high usability rating, but several participants cannot complete submission. Can the project be called successful on the rating alone?Reveal discussion

No. Investigate the task failures and the relationship between perceived usability and observed performance. Report both, revise the interface, and plan a further evaluation with an appropriate rationale.

Step 3 · Apply it to your own project

My research decision brief

Write short notes you can explain to a supervisor. This is a planning aid, not an assessed submission. Use fictional or non-identifying information. Notes remain in this browser; they are not sent to a tutor or shared with an account.

Who or what, which setting, and what kind of answer do you need?

Notes: __________________________________________________________ ________________________________________________________________

What can you access? What permissions and skills are still needed?

Notes: __________________________________________________________ ________________________________________________________________

Name a preferred approach and an alternative. Separate design, collection and analysis.

Notes: __________________________________________________________ ________________________________________________________________

Why does the preferred approach fit? Which tradeoff are you accepting?

Notes: __________________________________________________________ ________________________________________________________________

Record author, year, title, link and the specific claim each source supports.

Notes: __________________________________________________________ ________________________________________________________________

What assumptions, limitations and ethical or access issues must be addressed?

Notes: __________________________________________________________ ________________________________________________________________

What will you check next, with whom and by when?

Notes: __________________________________________________________ ________________________________________________________________

Loading saved progress… Download a copy before using a different device or clearing browser data.

Teaching and supervision resources

A seminar you can run in 50 minutes.

Adapt the question to your discipline and level. Students need access to Atlas and a way to record notes. Use Method library if the map is unsuitable for the room, device or access needs. For offline teaching, distribute the printed booklet and entry readings you are permitted to share.

Without screens, give each pair the scenario, two permitted entry readings and a blank brief. Replace Finder and Compare clicks with a paper table of purpose, evidence, requirements and limitations.

Learning outcome: By the end, learners should be able to locate relevant entries, distinguish decision layers, justify a provisional choice, identify a limitation and name a source and next action.

TimeActivityEvidence of learning
0 to 5 minPresent one fictional question. Ask what evidence would answer it.Each learner writes a purpose and one access constraint.
5 to 12 minDemonstrate Method library, one four-tab profile and its Sources tab.Learners locate minimum data requirements and one limitation.
12 to 25 minPairs try a worked scenario and inspect two candidates.Pairs identify whether entries are alternatives or complements.
25 to 35 minUse Compare and challenge a preferred option.Pairs explain one tradeoff and a condition that would change the choice.
35 to 45 minComplete the worksheet and inspect a source. If full reading is not feasible, record it as a next action.A provisional rationale distinguishes a source actually read from one still to read.
45 to 50 minShare an exit statement: choice, reason, limitation, next action.The tutor identifies unresolved assumptions to address next session.

Adapt by level

Undergraduate: Supply a bounded question and two candidate entries. Focus on naming the decision layer and one tradeoff.

Postgraduate: Ask learners to use their own question, challenge access assumptions and justify an alternative using literature.

Doctoral or staff development: Examine epistemological commitments, disciplinary conventions and what the atlas leaves unresolved.

Give feedback on the reasoning

Look for fit between question and evidence, a clear distinction between design and technique, an honest limitation, accurate use of a source, and an achievable next step.

If a learner says “the finder told me”, ask: “Which requirement did you check? What would make that suggestion unsuitable?”

After class, ask learners to read one original reference and revise their brief. Lesson completion is self-reported and is not a grade.

When you need a reminder

Glossary and help

Words you will meet

Methodology
The reasoning and assumptions that justify how a study is designed and conducted.
Paradigm
A broad set of assumptions about reality, knowledge and enquiry.
Design
The overall plan linking the question, evidence and claims.
Method or technique
A procedure for generating or analysing evidence. Read the entry’s type and layer.
Sampling
Decisions about which people, cases, materials or observations are included.
Measurement
How a concept is represented or observed, and how that representation is evaluated.
Reflexivity
Examining how the researcher’s position, assumptions and actions influence the research.
Assumption
A condition or proposition an approach relies on; identify and assess it rather than taking it for granted.
Integration
The planned relationship between qualitative and quantitative strands in mixed methods research.
Claim boundary
The limits of what the evidence and design allow you to conclude.

When something gets in the way

No search results
Clear the search, choose Show filters and Reset all filters, then try a shorter term. Header Search searches the full catalogue.
The finder will not continue
Choose an option. On Evidence, select at least one available evidence type.
The map is difficult to use
Choose Method library, the named method list or the small-screen layer cards. You do not need the 3D map to use the profiles.
I cannot add a fourth comparison
Compare holds up to three entries. Remove one first.
My comparison disappeared
Comparison selections and Finder answers save in this browser when storage is available. Check the saving message. A different device, blocked storage or cleared site data may explain missing work. Keep a downloaded brief.
A source is behind a paywall
Check your institutional library or an available author/repository version. Do not assume a summary is the full paper.
My notes are missing
Notes are local to the browser and device where you wrote them. Private browsing, clearing site data or changing devices can remove access. Use the downloaded brief.
Keyboard navigation
Use Tab to reach controls and Enter or Space to activate buttons. Use arrow keys in tab groups; Escape closes the quick guide and search dialogs. The Method library offers a text route to every entry.