Research-integrity, Ethics, and Data-Protection Reminder
Table of Contents
- 1. Lesson Overview
- 2. Learning Outcomes
- 3. Before You Begin
- 4. Retrieval Practice
- 5. Opening Scenario
- 6. Core Concepts
- 7. RIGOUR Walkthrough
- 8. Research-Process Walkthrough
- 9. Worked Example
- 10. Guided Learning Activities
- 11. Exercises
- 12. Homework Problems
- 13. Mini-Project
- 14. Self-Check Quiz
- 15. Assessment Connection
- 16. Assignment 1 Journal Connection
- 17. Key Takeaways
- 18. Key Terms
- 19. Looking Ahead
- 20. Sources and Further Reading
1. Lesson Overview
This lesson opens Unit 13 by establishing the foundational distinction between a broad research topic, a research problem, a purpose, a question, a hypothesis, and a method. You will learn how to use generative AI as a collaborative brainstorming and planning assistant—not as a substitute for your own scholarly judgment. The lesson emphasizes that AI may generate candidate research questions, suggest methods, and raise ethical or feasibility concerns, but you remain responsible for evaluating, justifying, and refining every element of the research plan.
By the end of this lesson, you will be able to formulate a clear, focused, and ethically defensible research question; identify legitimate and inappropriate AI roles in planning; and recognise when specialist, institutional, or community authority is required. This prepares you for Lesson 13.2, where you will move from a research question to a verified literature discovery and mapping process.
Research that begins with a poorly defined question, an unsuitable method, or an unexamined ethical assumption is unlikely to produce trustworthy results. AI can help you explore possibilities, but only you—working with your disciplinary and institutional authorities—can determine whether a research plan is sound, significant, and responsible.
2. Lesson Learning Outcomes
By the end of this lesson, you will be able to:
- Define research topic, research problem, research purpose, research question, hypothesis, and research method.
- Distinguish a broad topic from a researchable problem and a research purpose from a research question.
- Explain why AI may propose research possibilities but should not determine the final scholarly purpose or question.
- Evaluate a proposed research question for clarity, scope, significance, feasibility, evidence needs, assumptions, and ethical implications.
- Identify legitimate and inappropriate AI roles in research planning.
- Recognise required institutional, disciplinary, professional, legal, ethical, and community authority.
- Design a bounded, human-governed AI-supported research-planning workflow.
3. Before You Begin
- Assigned reading: Read the opening sections of Chapter 14 (Research and Scholarly Practice) up to the section on research questions. Review Chapter 7 (TRACE) and Chapter 8 (CLAIM) for verification and error-identification skills.
- Prerequisite knowledge: You should be able to apply PCAFE-V, SCOPE, PARTNER, FAIRER, PROTECT, CREATE, TRACE, and CLAIM. If you need a refresher, revisit Unit 10 (TRACE) and Unit 11 (CLAIM).
- Library and public-search options: You may use your institution's library, open-access repositories (e.g., Google Scholar, arXiv, SSRN), or approved AI tools if they are available to you. No activity requires a paid database or a specific AI system.
- Optional AI tool: If you have access to an approved generative AI system, you may use it for brainstorming and search-term generation. Do not upload any real participant data, confidential records, or unpublished work.
- Non-platform alternative: Every activity includes a fully supplied, synthetic alternative so that you are not disadvantaged if you choose not to use an AI system or paid database.
- Privacy and confidentiality: Use only synthetic, public, supplied, or explicitly authorised information.
- Research-ethics reminder: AI-assisted research planning does not replace ethics approval. Always consult your institution's research ethics board or equivalent authority.
- Copyright and source-use: Respect intellectual property and community-governed knowledge. Do not treat public visibility as permission to extract or reuse.
- Accessibility: All diagrams have text descriptions. If you need further accommodations, contact your accessibility office.
- Preparation time: Approximately 15–20 minutes for the reading and retrieval practice.
4. Retrieval Practice
These questions activate knowledge from Units 10–12. Attempt each one before revealing the feedback.
Question 1: A student asks an AI system to "find sources that support my argument that generative AI improves student learning." The AI returns a list of ten articles. Using TRACE, what is the first step the student should take?
Show answer
Answer: The student should trace each source by verifying that it exists, that the citation is accurate, and that the content actually supports the claim. TRACE (from Unit 10) requires identifying hallucinations, errors, omissions, invalid reasoning, and uncertainty. The student must not accept the list as verified.
Explanation: AI-generated source lists often contain fabricated references, misattributions, or selectively chosen works. Verification is an independent human responsibility.
Question 2: A researcher uses AI to summarise a 30-page government report. Using CLAIM (Unit 11), what must the researcher do before using that summary?
Show answer
Answer: The researcher must verify the summary against the original report. CLAIM requires checking that claims, evidence, citations, and interpretations are accurately represented. The researcher should read the original report (or substantial portions) and confirm that the AI summary does not omit, distort, or invent findings.
Explanation: AI summaries can lose context, flatten nuance, and introduce errors. Verification is a precondition for using any AI-generated summary in scholarly work.
Question 3: Using LEARN (Unit 12), explain the difference between using AI to generate flashcards for a topic and using AI to answer exam questions about that topic. Why does this distinction matter for research planning?
Show answer
Answer: LEARN distinguishes task completion from learning. Generating flashcards is a supportive task that can aid recall, provided the student verifies the content. Having AI answer exam questions substitutes for the student's own learning and undermines the development of disciplinary knowledge. In research planning, the distinction matters because AI may support brainstorming and organisation but should not replace the researcher's own understanding of the field, methods, or ethical considerations.
Explanation: Research requires deep, contextual understanding. AI-generated outputs are tools, not substitutes for scholarly expertise.
Research is not primarily about personal skill development; it is about generating new, verifiable, and ethically justified knowledge that can be shared and scrutinised by others. Research must meet disciplinary standards, ethics requirements, and community expectations — standards that go beyond individual learning.
5. Opening Scenario
Scenario: Dr. A. Rivera is a mid-career professional in a public-health agency. She has no formal research training but wants to conduct a small study to understand whether a new AI chatbot could improve patient adherence to medication regimens in a low-income community. She asks an AI system:
"Develop an important research study about generative AI."
The AI returns a proposal that includes:
- Broad topic: "AI in healthcare."
- Question: "What is the impact of AI chatbots on patient outcomes globally?"
- Assumption: "All communities have equal access to the internet and smartphones."
- Method: "Collect survey responses from 10,000 patients worldwide using a standardised online instrument."
- Ethics: No mention of consent, privacy, or community authority.
- Significance claim: "This study will revolutionise global healthcare."
Task: Diagnose what must be revised in this proposal. Identify at least six specific problems, and explain why each matters.
Show guided analysis
Diagnosis (six problems):
- Topic too broad: "AI in healthcare" is not a researchable problem. It covers an immense range of technologies, populations, outcomes, and contexts.
- Question not focused: "What is the impact … globally?" is overextended. It lacks a defined population, intervention, comparison, outcome, and time frame.
- Unsupported assumption: Assuming universal access to the internet and smartphones is empirically false and ethically problematic—it would exclude or misrepresent the very community Dr. Rivera wants to serve.
- Method unsuitable: An online survey of 10,000 patients worldwide is not feasible for a small, unfunded study. It also ignores cultural, linguistic, and literacy differences.
- Ethics omitted: No mention of research ethics review, informed consent, data privacy, confidentiality, or community engagement. The low-income community has governance rights that must be respected.
- Unverified significance claim: The claim that the study will "revolutionise global healthcare" is unsupported. Significance must be demonstrated through evidence and stakeholder needs, not asserted by an AI.
Why this matters: A research plan that is not grounded in a specific problem, a feasible method, and ethical authority cannot produce trustworthy knowledge. AI can help generate ideas, but the researcher must critically evaluate and revise every element.
6. Core Concepts
6.1 Research Topic
A research topic identifies a broad subject area. It is a starting point, not a destination. Topics are often stated in one or two words: generative AI in higher education; AI-supported public communication; automated software testing; AI and organisational knowledge; AI-supported environmental monitoring.
Common misconception: A topic is not yet a research question. Many students begin with a topic and mistakenly believe they have a research plan. The topic must be refined into a specific, researchable problem before any systematic inquiry can proceed.
6.2 Research Problem
A research problem identifies a specific gap, difficulty, uncertainty, inconsistency, or practical issue that warrants systematic investigation. A good research problem is:
- Specific enough to guide method selection;
- Grounded in existing literature, practice, or community need;
- Significant to stakeholders, the field, or the community;
- Feasible within available time, resources, and authority.
AI may propose problems, but the researcher must verify that the problem is genuine and not merely fabricated or exaggerated.
6.3 Research Purpose
The research purpose states what the study intends to investigate, describe, explain, compare, evaluate, design, or understand. It is a declarative statement that connects the problem to the intended contribution.
Example: "The purpose of this study is to describe how undergraduate students in a 300-level online course use generative AI to generate research questions, and to identify the verification strategies they employ."
The purpose should be legitimate, feasible, ethically defensible, and connected to evidence. AI may help you articulate a purpose, but you must evaluate whether it is appropriate for your context.
6.4 Research Question
A research question translates the purpose into an interrogative form that can be answered through systematic inquiry. A well-formed research question is:
- Clear: The key concepts are defined and understandable.
- Focused: It narrows the scope to a manageable, coherent inquiry.
- Researchable: It can be answered with evidence that is obtainable and ethically permissible.
- Methodologically answerable: The methods available can produce evidence that addresses the question.
- Ethically permissible: The research can be conducted without causing harm or violating rights.
- Significant: The answer would contribute to knowledge, practice, or community well-being.
AI may generate candidate questions, but you must evaluate and justify the final question. Do not adopt an AI-generated question without critical scrutiny.
6.5 Hypothesis
A hypothesis is a testable proposition about an expected relationship or difference. It is most common in quantitative and experimental research. Not every study requires a hypothesis; qualitative and exploratory studies often do not have a formal hypothesis.
Example: "Students who receive AI-generated feedback on their research questions will generate more focused questions than students who receive no feedback."
AI may suggest hypotheses, but you must ensure that the hypothesis is grounded in theory, prior evidence, and methodological appropriateness.
6.6 Research Method
A research method is the systematic approach used to collect and analyse evidence to answer the research question. Methods may be:
- Quantitative: Surveys, experiments, statistical analysis, computational modelling;
- Qualitative: Interviews, ethnography, case study, thematic analysis;
- Mixed methods: Combining quantitative and qualitative approaches;
- Design science: Building and evaluating artefacts (e.g., software, policies);
- Legal/doctrinal: Analysing statutes, case law, or regulations;
- Historical: Analysing archival or historical sources;
- Systematic review: Synthesising existing research evidence.
No single method fits every question. The method must align with the question, the evidence needed, and the ethical constraints.
6.7 Methodological Fit
Methodological fit refers to the coherence among the research question, the data, the method, the analysis, and the interpretation. When these elements are aligned, the study is more likely to produce valid, trustworthy conclusions.
AI-generated method recommendations must not substitute for disciplinary methodology. You are responsible for selecting a method that fits your question and context.
6.8 AI Roles in Research Planning
Generative AI may play supportive, bounded roles in research planning, such as:
- Brainstorming candidate research topics and questions;
- Generating subquestions and key concepts;
- Suggesting search terms and synonyms;
- Proposing possible methods and limitations;
- Raising stakeholder, risk, or ethics questions;
- Providing a provisional structure for a research plan.
AI should not be used for:
- Making autonomous ethics decisions;
- Determining the final research question without human evaluation;
- Selecting sources without verification;
- Interpreting qualitative or quantitative data as a final authority;
- Assuming authorship or scholarly accountability.
Every AI proposal must be evaluated by the researcher and, where appropriate, by relevant authorities.
6.9 Summary: Topic → Problem → Purpose → Question → Method
| Element | Definition | Example |
|---|---|---|
| Topic | Broad subject area | Generative AI in higher education |
| Problem | Specific gap, difficulty, or uncertainty | Students use AI to generate research questions but may not verify the outputs; no study has examined verification practices. |
| Purpose | What the study intends to do | To describe how undergraduate students verify AI-generated research questions. |
| Question | Interrogative form of the purpose | What verification strategies do undergraduate students use when they generate research questions with AI? |
| Method | Approach to collect and analyse evidence | Qualitative: semi-structured interviews and analysis of AI interactions. |
Try this: Ask an AI system to generate a research question about a topic you are interested in. Then evaluate the question using the criteria above (clarity, focus, researchability, ethics, significance). Do not accept the question; instead, revise it until it meets the criteria. This is a safe, low-risk way to practice human-governed AI-assisted planning.
7. RIGOUR Walkthrough
The RIGOUR framework is the author-approved model for responsible AI-supported research and scholarly practice. Because the exact expansion has not been supplied in this material, we use the operational definition provided by the course documents. When the exact expansion is supplied, replace this placeholder with the authorised terms.
RIGOUR addresses the following research governance requirements:
- R – Research purpose and question that are legitimate, focused, and significant.
- I – Information, data, source, and ethics governance that respects authority, consent, and community rights.
- G – Method selection and methodological fit that align question, data, and analysis.
- O – Verification of analysis and interpretation through independent checks and human judgment.
- U – Documentation, reproducibility, and provenance that allow reconstruction and audit.
- R – Uncertainty, disclosure, human responsibility, and review that make limitations and accountability explicit.
Note: [INSERT AUTHOR-APPROVED RIGOUR EXPANSION] when supplied.
In Lesson 13.1, we focus on the first element (Research purpose and question) and preview how it connects to the others. Later lessons will address information governance, method, verification, reproducibility, and disclosure in detail.
7.1 RIGOUR Element 1: Research Purpose and Question
A research purpose and question that are legitimate, clearly defined, appropriately scoped, methodologically answerable, ethically permissible, and significant within the relevant context.
Guiding questions:
- What is the specific problem this research addresses?
- What is the intended contribution of this research?
- Is the question focused enough to be answered with the available evidence and methods?
- What ethical, cultural, or community considerations are embedded in the question?
Interdisciplinary example: A team in environmental studies wants to study how local communities perceive AI-based flood prediction systems. The team refines a broad topic ("AI and climate") into a focused question: "How do residents of the flood-prone coastal community of X perceive the trustworthiness of AI-based flood alerts?"
Common misuse: Accepting an AI-generated question such as "What is the impact of AI on the environment?" without narrowing the population, geography, time frame, or evidence type.
Verification requirement: The researcher must demonstrate that the question is grounded in existing evidence (e.g., community needs, prior literature, official reports) and that it does not assume what it seeks to prove.
Escalation condition: If the question touches on sensitive populations, cultural knowledge, or regulated data, escalate to the research ethics board, community authority, or other relevant body before proceeding.
Residual limitation: A well-formed question does not guarantee that the research is ethical, feasible, or significant; it is a necessary but insufficient condition for responsible research.
Show application exercise: RIGOUR question audit
Prompt: Evaluate the following research question using the RIGOUR "purpose and question" criteria: "How does AI affect education?"
Sample analysis:
- Clarity: Low – "AI" and "education" are undefined.
- Focus: Low – no specified population, level, context, or outcome.
- Researchability: Low – the question cannot be answered with a single study.
- Significance: Moderate – education is important, but the question is too vague to indicate a contribution.
- Ethics: Undefined – no mention of participants, consent, or data.
Revised version: "How do undergraduate engineering students at a mid-sized Canadian university use generative AI tools to complete design assignments, and what verification strategies do they report?"
8. Research-Process Walkthrough: Topic to Question to Planning
This walkthrough illustrates how a researcher moves from a broad topic to a focused research question and plan, with bounded AI support at each stage.
Stage 1: Identify a topic
Human activity: The researcher notes a personal or professional interest. Example: "I'm interested in how generative AI is used in journalism."
Bounded AI activity: The researcher may ask an AI system: "What are some current topics in AI and journalism?" The AI generates a list (e.g., automated news writing, fact-checking, source verification, audience engagement).
Verification/Authority: The researcher verifies that these topics are current by scanning recent journalism-related publications and professional forums.
Stage 2: Define a research problem
Human activity: The researcher identifies a specific gap or difficulty. Example: "While AI is used to draft news articles, there is little research on how journalists verify AI-generated content before publication."
Bounded AI activity: The researcher may ask: "What are the main verification challenges journalists face with AI-generated content?" The AI suggests categories (e.g., source tracing, factual accuracy, bias detection).
Verification/Authority: The researcher checks professional journalism standards (e.g., codes of ethics, industry reports) to confirm that these challenges are legitimate and not invented.
Stage 3: Articulate a purpose
Human activity: The researcher writes a purpose statement. Example: "The purpose of this study is to describe how journalists verify AI-generated content in their daily work and to identify the verification strategies they use."
Bounded AI activity: The researcher may ask: "What are some alternative ways to state this purpose?" The AI offers variations, which the researcher evaluates for clarity and fit.
Verification/Authority: The researcher ensures the purpose is achievable within the project scope and aligns with disciplinary expectations.
Stage 4: Formulate a research question
Human activity: The researcher drafts a question. Example: "What verification strategies do journalists report when they use generative AI to draft news articles?"
Bounded AI activity: The researcher may ask: "Generate three alternative phrasings of this question." The AI provides options, which the researcher critically compares.
Verification/Authority: The researcher evaluates the question for clarity, focus, researchability, ethics, and significance. She may consult a supervisor or librarian.
Stage 5: Select a method
Human activity: The researcher chooses a qualitative interview approach, as the goal is to understand journalists' experiences and strategies.
Bounded AI activity: The researcher may ask: "What are the strengths and limitations of using semi-structured interviews for this research?" The AI summarises common considerations, but the researcher must evaluate them against the specific context.
Verification/Authority: The researcher consults qualitative methods literature and may seek input from an experienced interviewer.
Stage 6: Identify ethics and authority
Human activity: The researcher identifies that journalists are human participants; therefore, research ethics board approval is required. She also considers whether any information could be sensitive (e.g., unpublished stories, confidential sources).
Bounded AI activity: The researcher may ask: "What ethics considerations are typical for interview studies in journalism?" The AI provides a list, but the researcher must confirm with her institution's ethics board.
Verification/Authority: The researcher submits an ethics application and receives approval before beginning any data collection.
Stage 7: Define AI role boundaries
Human activity: The researcher decides that AI may be used for brainstorming, search-term generation, and organising interview transcripts, but not for interpreting the interviews or drawing conclusions.
Bounded AI activity: The researcher documents which AI system, prompts, and outputs were used, and retains all records.
Verification/Authority: The researcher ensures that the AI use complies with ethics approval and data-management policies.
At every stage, the researcher exercises judgment and authority. AI is an assistant, not a decision-maker. The researcher verifies, evaluates, and justifies every element of the plan.
9. Worked Example
This worked example shows how a researcher applies the concepts from this lesson to a synthetic, low-risk scenario.
Scenario: Maya is a graduate student in public health. She is interested in whether community health workers in a rural region find AI-powered translation tools useful when communicating with patients who speak a different language. She has no funding and limited time. She decides to use AI only for brainstorming and organisation.
Step 1: Topic → Problem
Topic: AI in community health.
Problem: Health workers in rural areas often face language barriers with patients. AI translation tools exist, but little is known about how health workers perceive their accuracy and usefulness in real-world settings.
Justification: Maya searches Google Scholar and finds only two relevant studies, both from urban settings. The problem is genuine and under-researched.
Step 2: Purpose → Question
Purpose: To describe how community health workers in a rural region perceive and use AI-powered translation tools during patient interactions.
Research question: "How do community health workers in X rural region perceive the usefulness and accuracy of AI-powered translation tools in their daily patient interactions?"
Evaluation:
- Clear? Yes – the key concepts are defined.
- Focused? Yes – a specific region, population, and tool type.
- Researchable? Yes – interviews with health workers are feasible.
- Ethically permissible? Yes – with ethics review and informed consent.
- Significant? Moderate – could inform training and tool selection.
Step 3: Methodological fit
Method: Qualitative – semi-structured interviews with 8–10 health workers. Thematic analysis of transcripts.
Fit: The question asks about perceptions and experiences, which are best explored qualitatively. The sample size is manageable for a small project.
Step 4: AI role and boundaries
- Permitted AI support: Brainstorming interview questions, generating search terms for literature, organising interview transcripts.
- Prohibited AI involvement: Interpreting transcripts, drawing conclusions, or making recommendations.
- Human responsibility: Maya will conduct all interviews, code the transcripts manually, and write the final report.
Step 5: Ethics and authority
- Maya submits an ethics application to her institution's research ethics board.
- She obtains informed consent from all participants.
- She does not upload interview transcripts to any AI system; all data remain on a secure institutional drive.
- She respects the confidentiality of health workers and the patients they serve.
Step 6: Documentation and disclosure
Maya documents:
- All AI prompts and outputs (dates, system names, versions where available);
- Her human decisions and revisions;
- Ethics approval number and consent procedures;
- A clear AI-use disclosure in her final report.
Maya used AI only for brainstorming, search-term generation, and organisational support. She made all substantive decisions herself and verified every AI output. Her research plan is bounded, ethically considered, and methodologically coherent.
10. Guided Learning Activities
Instructions: Classify each statement as a topic, problem, purpose, question, or method. Then explain your reasoning.
- "This study aims to explore how librarians use AI to help students with literature searching."
- "Generative AI and academic libraries."
- "We will conduct semi-structured interviews with 12 librarians."
- "Librarians are increasingly asked to support AI-assisted research, but we do not know what training or resources they need."
- "What training do academic librarians report needing to support students' use of AI in literature searching?"
Show classification and explanation
- Purpose – states what the study intends to do.
- Topic – a broad subject area.
- Method – describes the data-collection approach.
- Problem – identifies a gap.
- Question – an interrogative form of the purpose.
Explanation: The sequence shows refinement from a broad topic (2) to a specific problem (4), a purpose (1), a method (3), and a focused research question (5).
Instructions: Evaluate the following candidate question using the criteria in Section 6.4. Identify at least three weaknesses and suggest a revision.
"Does AI help students learn?"
Show evaluation and revised question
Weaknesses:
- Clarity: "AI" and "learn" are undefined.
- Focus: The question is enormous—it could refer to any level, subject, or context.
- Researchability: It is not a single answerable question.
- Assumptions: It assumes "help" is binary.
Revised question: "How do undergraduate students in a first-year composition course use a generative AI writing assistant to revise their drafts, and what revisions do they report making?"
Instructions: For each proposed AI role, classify it as appropriate bounded support, potentially appropriate with safeguards, inappropriate substitution, specialist review required, or insufficient information.
- An AI system generates 50 possible research questions on a topic.
- An AI system selects the "best" research question and method without human review.
- An AI system suggests a qualitative method for a study that aims to understand lived experiences.
- An AI system determines that the research does not require ethics approval.
- An AI system helps the researcher organise notes and identify potential themes.
Show classifications and explanations
- Appropriate bounded support – brainstorming possibilities is a legitimate AI role.
- Inappropriate substitution – selection and justification must remain human.
- Appropriate bounded support – suggesting methods can be helpful, but the researcher must verify fit.
- Inappropriate substitution – AI cannot determine ethics approval.
- Appropriate bounded support – organisation and provisional theme generation are within scope.
11. Exercises
Complete these seven exercises. Each includes a solution or solution design inside a details element.
Task: For each of the following six statements, identify whether it is a topic, problem, purpose, question, hypothesis, or method. Then arrange them in a logical order from most general to most specific.
- A. "This research will compare the effectiveness of two AI-powered translation tools in a rural health clinic."
- B. "AI translation tools in healthcare."
- C. "Health workers in rural clinics may lack consistent access to translation services, and the accuracy of AI tools in this context is unknown."
- D. "We will use a quasi-experimental design with pre- and post-test surveys."
- E. "Health workers who use Tool A will report higher satisfaction than those who use Tool B."
- F. "What are the comparative satisfaction rates of health workers using Tool A versus Tool B in a rural clinic?"
Show solution
Classification: B – Topic; C – Problem; A – Purpose; F – Question; E – Hypothesis; D – Method.
Logical order: B → C → A → F → E → D
Task: The following research question is too broad. Rewrite it to be focused, researchable, and ethically defensible. "What is the impact of generative AI on society?" Specify population, context, outcome, and evidence type.
Show sample revision
Revised question: "How do urban planning professionals in Canada perceive the impact of generative AI on public engagement processes, and what concerns do they report about bias and transparency?"
Task: A researcher proposes: "We will survey 500 university students about their use of AI for coursework. All students have reliable internet access and smartphones." Identify at least three hidden assumptions. Why is each assumption problematic?
Show analysis
- All students have reliable internet and smartphones. Not true for many students; would bias the sample.
- All students use AI for coursework in similar ways. Use likely varies by discipline, year, and preference.
- Survey responses accurately reflect actual behaviour. Students may over- or under-report AI use.
Task: A researcher asks: "What are the lived experiences of first-generation university students using AI for academic writing?" The researcher proposes a quantitative survey with closed-ended questions. Is this a good methodological fit? Explain why or why not, and suggest an alternative if needed.
Show analysis
Assessment: Poor fit. "Lived experiences" implies depth, context, and meaning, which are better explored qualitatively.
Alternative: Semi-structured interviews or focus groups.
Task: A researcher asks an AI system: "What research questions should I ask about AI and mental health?" The AI returns ten questions. The researcher selects the first question and proceeds without further evaluation. Classify this AI role and explain whether it is appropriate.
Show classification
Classification: Inappropriate substitution. The AI generated possibilities (appropriate), but the researcher delegated selection without evaluation (inappropriate).
Task: A researcher plans to study the experiences of Indigenous youth with AI-based mental health apps. The researcher intends to recruit participants through social media and use an AI system to analyse interview transcripts. Identify at least four ethics or authority concerns.
Show analysis
- Indigenous data governance: Engage the relevant Indigenous community or governance body.
- AI analysis of transcripts: Uploading transcripts may violate confidentiality.
- Mental health content: May require additional ethics review and support.
- Social media recruitment: Raises privacy and consent questions.
Task: For a research question of your choice (or use the supplied one), design a workflow that specifies: which activities are human-only; which activities may involve AI, and what boundaries apply; how AI outputs will be verified; what authority or ethics approval is required.
Supplied question: "How do public-sector managers perceive the usefulness of generative AI for policy analysis?"
Show solution design
- Human-only activities: Define problem, purpose, final question, method; obtain ethics approval; conduct interviews; interpret findings; write report.
- AI-supported activities (bounded): Brainstorm interview questions; generate search terms; organise transcripts (no uploads).
- Verification: All AI materials reviewed and revised; literature discovered by AI verified against originals.
- Authority/ethics: Ethics approval; permission from public-sector organisations; informed consent.
12. Homework Problems
Show sample answer
Research problem: A specific gap, difficulty, uncertainty, or practical issue that warrants investigation.
Research purpose: A statement of what the study intends to do.
Difference from topic: A topic is a broad subject area; a problem identifies a specific gap; a purpose states the action to address the problem.
Show sample answer
Research question: An interrogative statement that the research aims to answer.
Hypothesis: A testable proposition about an expected relationship or difference.
When not required: In exploratory or descriptive research (e.g., qualitative studies), there may be no formal hypothesis.
Show sample revision
Revised: "How do supply-chain managers in Canadian manufacturing firms perceive the impact of generative AI on inventory forecasting accuracy, and what implementation challenges do they report?"
Show evaluation
Clarity: Low – "best" is undefined.
Scope: Overextended.
Significance: Moderate – too vague.
Feasibility: Low.
Ethics: Undefined.
Show analysis
Why insufficient: "Hot topic" is a popularity claim, not an evidence-based justification for significance.
Stronger justification: Ground significance in gaps in knowledge, stakeholder needs, or practical problems.
Show analysis
- Logistical: Reaching 5,000 participants without internet would require in-person contact.
- Cultural/linguistic: Survey instruments would need translation and cultural adaptation.
Show analysis
Fit: Poor. A correlational survey cannot establish causal effects. Causal questions require experimental or quasi-experimental designs.
Show analysis
Fit: Poor. "Lived experiences" implies depth and meaning, which are best explored qualitatively.
Show classification
Classification: Inappropriate substitution. The AI suggested a method, but the researcher did not evaluate whether the method fits the question.
Show analysis
- Public ≠ consent.
- Privacy: Tweets may contain identifiable information.
- Institutional authority: Many institutions require ethics review.
- AI use: Uploading tweets may violate platform terms and data laws.
Show analysis
Problem: Nurses are key stakeholders. Without their input, the research may miss practical constraints, usability issues, or ethical concerns.
What to do: Engage nurses in the research design phase.
Show analysis
Boundary crossed: The researcher used AI to generate substantive research content without independent verification. The researcher must verify all sources, develop the method themselves, and consult the ethics board.
Show sample response
Example (education): If I ask "What are the experiences of first-year students with AI tutoring systems?", a quantitative survey would be a poor fit because experiences are qualitative. Interviews or reflective journals would be better.
13. Mini-Project: Responsible Research Planning Brief
Purpose: To integrate the concepts from Lesson 13.1 into a bounded, low-risk research-planning document.
Learning outcomes: Formulate a focused research question; evaluate its feasibility, significance, and ethics; define appropriate AI roles; and identify required authority.
Estimated effort: 90–120 minutes.
Available contexts: Choose one of the following synthetic scenarios, or use a low-risk topic of your own (subject to instructor approval).
- A. Education: How do students in a 300-level course use AI to generate research questions?
- B. Business: How do small-business owners perceive the usefulness of AI marketing tools?
- C. Health: What are the experiences of caregivers using AI monitoring devices for elderly relatives?
- D. Public administration: How do government communicators use AI to draft public-facing materials?
- E. Humanities: How do undergraduate history students use AI for primary-source analysis?
Deliverable: A structured planning brief (approximately 1,000–1,200 words or equivalent).
Required sections
- Broad topic
- Research problem
- Evidence that the problem warrants investigation
- Research purpose
- Candidate research question
- Subquestions
- Hypothesis (if appropriate)
- Proposed method
- Methodological-fit explanation
- Intended evidence
- Stakeholders
- Proposed AI role
- Permitted AI support
- Prohibited AI involvement
- Information and data boundaries
- Ethics or authority questions
- Community-governance questions
- Feasibility
- Limitations
- Escalation condition
- Stop condition
- Preliminary recommendation
Show staged hints
Hint 1: Start with a specific gap. What is not known or not understood?
Hint 2: Check your question against the criteria: clarity, focus, researchability, ethics, significance.
Hint 3: For each AI task, ask: "Could this task be done without AI?" If the answer is yes, AI is a support tool.
- I have defined a specific research problem, not just a topic.
- My research question is clear, focused, researchable, and ethically permissible.
- I have justified why the problem matters.
- My proposed method fits the question.
- I have identified appropriate and inappropriate AI roles.
- I have considered ethics and authority requirements.
- I have addressed feasibility and limitations.
- I have included escalation and stop conditions.
Show partial example: Education context
Topic: AI in graduate supervision.
Problem: Graduate supervisors lack evidence-based guidance on how to discuss AI use with their students.
Purpose: To explore supervisors' current practices and information needs.
Question: "What guidance do graduate supervisors report needing to support students' appropriate use of generative AI in thesis research?"
Method: Qualitative – semi-structured interviews with 10 supervisors.
AI role: Brainstorm interview questions; generate search terms; organise interview notes. Prohibited: Interpreting interviews, drawing conclusions.
Ethics: Ethics approval required. All participants must give informed consent.
14. Self-Check Quiz
Question 1. Which of the following is a research problem (not a topic, purpose, or question)?
Show answer and explanation
Answer: C.
Explanation: A is a topic, B is a purpose, D is a question. C identifies a specific gap.
Question 2. A researcher asks AI to "generate a research question about AI and sustainability." The AI returns: "What is the impact of AI on sustainability?" Which of the following best describes this question?
Show answer and explanation
Answer: B. Overbroad and unfocused.
Question 3. Which of the following is an appropriate AI role in research planning?
Show answer and explanation
Answer: C. Brainstorming candidate research questions for human evaluation.
Question 4. What is methodological fit?
Show answer and explanation
Answer: B.
Question 5. A researcher states: "My study is significant because it uses AI." Why is this not a sufficient justification?
Show answer and explanation
Answer: B.
Question 6. Which of the following is an inappropriate AI role in research planning?
Show answer and explanation
Answer: B.
Question 7. What is the purpose of an escalation condition in a research plan?
Show answer and explanation
Answer: B.
Question 8. Which of the following statements about research purpose is true?
Show answer and explanation
Answer: B.
15. Assessment Connection
- Assignment 1 (AI Learning and Reflection Journal): Preserve evidence of your research-question development, AI role definitions, and ethics-authority mapping.
- Assignment 5 (Verification and Fact-Checking): The skills you practised in Unit 11 (CLAIM) apply directly to verifying AI-generated research plans and questions.
- Assignment 4 (Applied AI Workflow Project): The research-planning workflow you design in this lesson can inform your workflow design project.
- Capstone (Assignment 6): The ability to formulate a focused, ethically defensible research question is foundational.
16. Assignment 1 Journal Connection
- Prompt 1: How does a research topic differ from a research problem and a research question?
- Prompt 2: In what ways did AI help you refine your research question? In what ways did it potentially distort or limit your thinking?
- Prompt 3: What ethics or authority questions did you encounter when planning your research?
- Prompt 4: How has your understanding of "responsible AI use in research" changed?
Show reflection planning checklist
- I can distinguish topic, problem, purpose, question, and method.
- I can evaluate a question for clarity, focus, feasibility, ethics, and significance.
- I can identify appropriate and inappropriate AI roles in planning.
- I can identify ethics and authority requirements.
- I have documented my AI interactions and human decisions.
17. Key Takeaways
- A research topic is not a research problem.
- A research purpose states what the study intends to do; a research question translates that purpose into an interrogative form.
- A well-formed research question is clear, focused, researchable, ethically permissible, and significant.
- Methodological fit means that the research question, data, method, analysis, and interpretation are coherent.
- AI may support planning through brainstorming, search-term generation, and organisation, but it must not determine the final research question or assume scholarly accountability.
- Research ethics and authority are non-negotiable.
- Document every AI interaction and human decision.
18. Key Terms
- Research topic
- A broad subject area that serves as a starting point for inquiry.
- Research problem
- A specific gap, difficulty, uncertainty, inconsistency, or practical issue that warrants systematic investigation.
- Research purpose
- A statement of what the study intends to investigate, describe, explain, compare, evaluate, design, or understand.
- Research question
- A clear, focused, researchable, ethically permissible, and significant interrogative that guides the study.
- Hypothesis
- A testable proposition about an expected relationship or difference, used when appropriate.
- Research method
- A systematic approach to collecting and analysing evidence to answer the research question.
- Methodological fit
- The coherence among research question, data, method, analysis, and interpretation.
- AI-supported research planning
- Using generative AI for bounded support tasks such as brainstorming, search-term generation, and organisation, while preserving human judgment.
- Research ethics
- The principles and procedures that govern the conduct of research.
- Data governance
- The policies, authority, and practices that control the collection, use, storage, sharing, and retention of data.
- Community governance
- Collective authority over data, knowledge, language, records, interpretations, or cultural materials held by a community or Indigenous group.
- Escalation condition
- A trigger that requires consultation with specialist, institutional, or community authority.
- Stop condition
- A situation that requires pausing or discontinuing research due to risk, uncertainty, or lack of authority.
- Provenance
- A documented record of the origin, transformation, and decisions affecting research questions, sources, data, prompts, code, outputs, and interpretations.
19. Looking Ahead
Next lesson: Lesson 13.2: Literature Discovery, Mapping, and Scholarly Synthesis – You will learn how to use AI cautiously for search-term development and literature discovery, distinguish discovery from verification, build a literature map, and construct a scholarly synthesis.
Unit 13 continues: Lessons 13.3 and 13.4 will cover AI-assisted analysis, interpretation, research ethics, reproducibility, documentation, and the full RIGOUR framework.
After Unit 13: You will complete Module 4 and move to Module 5: AI for Professional Practice and Everyday Life.
20. Sources and Further Reading
- Primary: COMP 3270 Textbook, Chapter 14: Research and Scholarly Practice.
- Supporting: Chapter 7 (TRACE); Chapter 8 (CLAIM); Chapter 12 (LEARN); selected sections of Chapter 13.
- Appendix references: Appendix A (Glossary), Appendix C (Verification Toolkit), Appendix D (Responsible-Use Toolkit), Appendix E (AI Workflow Design Toolkit).
- Institutional guidance: Refer to your institution's research ethics board, privacy office, and data-management policies.
- Indigenous research: Consult applicable Indigenous research frameworks, community governance protocols, and the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (TCPS 2), Chapter 9.
- Further reading: [INSERT VERIFIED RESEARCH-METHODS REFERENCE].