Research Questions and Literature Search

The Research Question Problem

Most academic papers start with a question that’s either too big or too small.

Too big: “How does social media affect mental health?” — You could write ten books on this. Where do you even start?

Too small: “Did Twitter usage increase among 19-year-olds in Ohio in March 2024?” — Maybe interesting, but so narrow it’s hard to connect to anything bigger.

The sweet spot is a question that’s specific enough to investigate rigorously but significant enough that the answer matters to your field.

From Topic to Question

Most researchers start with a topic, not a question. AI helps you bridge the gap:

My general topic area is: [your topic]
My discipline is: [your field]

Help me:
1. Identify 5 specific research questions within this topic
2. For each, explain what gap in knowledge it addresses
3. Rate each on a 1-5 scale for feasibility (data availability, scope)
4. Suggest which one would make the strongest contribution to the field

The PICO Framework (Adapted)

Originally from medical research, this framework works across disciplines:

Population: Who or what are you studying?
Intervention/Variable: What factor are you examining?
Comparison: Against what?
Outcome: What are you measuring?

Example transformation:

Vague: “How does remote work affect productivity?”
PICO: “How does fully remote work (I) compare to hybrid arrangements (C) in affecting project completion rates (O) among software development teams (P)?”

The second version is researchable. The first is a conversation topic.

Building Your Search Strategy

Once your question is framed, you need sources. AI helps you build a systematic search strategy:

My research question is: [your question]
My field is: [discipline]

Help me build a search strategy:
1. Key search terms and synonyms for each concept
2. Boolean search strings I can use in databases
3. Relevant academic databases for this topic
4. Types of sources I need (empirical studies, reviews, theoretical papers)
5. Date range that makes sense for this topic

Search String Construction

Good database searches use Boolean operators:

Operator	Purpose	Example
AND	Both terms must appear	“remote work” AND productivity
OR	Either term can appear	“remote work” OR “telecommuting”
NOT	Exclude a term	leadership NOT “thought leadership”
" "	Exact phrase	“organizational culture”
*	Wildcard	psycholog* (matches psychology, psychological)

AI generates these strings quickly, but you should understand the logic.

Quick Check

Look at these two search strategies and identify which one will produce better results:

A) Search Google for “social media mental health” and use the first 20 results B) Search PsycINFO for (“social media” OR “Instagram” OR “TikTok”) AND (“mental health” OR “wellbeing”) AND (“adolescents” OR “teenagers”) with a date filter of 2020-2026

See answer

Strategy B is far superior. It uses a specialized database, includes synonyms (OR), combines concepts (AND), specifies the population, and filters by date. Strategy A would return popular articles, not peer-reviewed research, with no specificity.

Evaluating Sources with AI

Not all papers deserve space in your literature review. Use AI to evaluate sources:

Here's the abstract of a paper I'm considering for my literature review:

[Paste abstract]

My research question is: [your question]

Evaluate this source:
1. How relevant is it to my specific question? (High/Medium/Low)
2. What methodology was used? Is it rigorous?
3. How recent is this research? Is the timeframe still relevant?
4. What does this paper contribute that I couldn't get elsewhere?
5. Should I include this in my review? Why or why not?

Critical rule: Always read the actual paper. AI evaluates abstracts, but the details, methodology, and nuance live in the full text.

Organizing Your Sources

As you collect sources, organize them thematically, not chronologically:

Theme-Based Organization

AI: I've found these 15 sources for my research on [topic].
Here are their titles and key findings:

[List sources]

Help me organize these into thematic clusters:
1. Group sources by the subtopics they address
2. Identify where sources agree and disagree
3. Flag any gaps—subtopics with few or no sources
4. Suggest a logical order for presenting these themes

The Source Matrix

Create a spreadsheet with these columns:

Source	Theme	Method	Key Finding	Agrees With	Disagrees With	Quality
Author, Year	Category	Qual/Quant	Main result	Related sources	Contradicting sources	Strong/Moderate/Weak

This matrix becomes the foundation of your literature review in Lesson 3.

Managing Citation Information

Every source you read should have complete citation information saved immediately. Don’t wait.

Use your reference manager (Zotero, Mendeley, EndNote) to:

Import citation data directly from databases
Store PDFs alongside citations
Add your notes and tags
Generate formatted bibliographies later

Use AI to verify citation completeness:

Check this citation for completeness in APA 7th edition format:
[Paste your citation]

Is anything missing? Is the format correct?

Exercise: Build Your Research Foundation

Write your research question using the PICO framework
Generate a search strategy with AI (terms, databases, Boolean strings)
Search one database and collect 10 potentially relevant sources
Evaluate each source’s relevance with AI assistance
Organize your sources into thematic clusters

This foundation makes the literature review (next lesson) dramatically easier.

Key Takeaways

Research questions must be specific enough to investigate but significant enough to matter
Use the PICO framework to transform vague topics into researchable questions
Build systematic search strategies with Boolean operators and database-specific terms
Always verify AI-suggested references actually exist in real databases
Organize sources thematically, not chronologically, using a source matrix
Save complete citation information immediately—don’t create work for yourself later

Up next: In the next lesson, we’ll dive into Synthesizing Literature Reviews.