50+ Research Title Examples & Formulas for Every Discipline

You’re staring at a half-finished paper, and the title still sounds like a folder name: “Climate Change Research Paper” or “Social Media Study.” Bad titles do real damage. They make a specific paper look generic.

A strong research title usually follows one of six patterns: question, finding, comparison, problem-solution, descriptive, or mechanism. Pick the pattern that matches your paper, put the searchable keywords near the front, and cut anything that doesn’t help a reader predict the study.

This guide gives you 50+ examples across STEM, medicine, psychology, education, law, business, history, literature, and other fields. If you’re already collecting PDFs and notes, an AI research workspace like Otio can help you generate title variants from your actual sources instead of asking a blank chatbot to guess your field.

Who this list is for

This is for the student or researcher who already has a topic but can’t make the title behave.

Undergraduates usually need the most help narrowing scope. “The effects of social media” feels safe because it covers everything; it fails because it says almost nothing. A stronger title names the platform, population, outcome, and method if the method matters.

Graduate students and PhD candidates have the opposite problem. The paper is specific, but the title gets overloaded with every variable, dataset, theory, and caveat. That’s how you get a 31-word title that nobody can remember.

Researchers in STEM, humanities, social sciences, and professional fields also face different norms. A medicine title may foreground the trial design. A history title may foreground the archive. A law title may foreground the doctrine or remedy. The same formula won’t travel cleanly across all of them.

If you’re still choosing the paper’s direction, start with research topic examples or research questions examples first. Titles work best after the question has stopped wobbling.

How we picked these 50+ examples

The examples below are built from common title patterns in published research, not from random “academic-sounding” phrasing. Research article titles follow discipline conventions: medicine, law, computer science, history, and education don’t signal value the same way.

Large title studies back this up. A ScienceDirect paper examining 5,070 titles across six disciplines found that titles vary by structure and content focus across human and physical sciences, especially in how they balance information and appeal (ScienceDirect’s study of research article titles).

We grouped titles by structure first: question-based, hypothesis-driven, comparative, problem-solution, descriptive, and mechanistic. Then we mapped those structures onto disciplines where they naturally appear.

You’ll see a mix of realistic published-style titles and AI-assisted variations. The goal isn’t to pretend every example is a famous paper. The goal is more useful: give you repeatable title shapes you can adapt without sounding like a template generator.

There’s a catch. AI-generated titles often sound polished before they’re accurate. A Nature portfolio study comparing human-authored and ChatGPT-generated medical research titles analyzed 300 human titles and 300 ChatGPT-generated titles, looking at length, form, syntax, and content focus (Nature Humanities and Social Sciences Communications on AI-generated research titles). That’s the useful frame: AI can draft options, but the researcher has to decide what the paper actually claims.

What makes a strong research title

A strong title does three jobs at once. It tells the reader the subject, signals the method or evidence when relevant, and puts the right keywords where search systems can find them.

Specificity beats breadth. “The role of dopamine in reward processing” beats “Neurotransmitters and behavior” because it names the mechanism and process. Even better: “Dopamine prediction errors shape reward learning in adolescents: Evidence from fMRI.”

Keywords should come early. If the paper is about maternal mortality, don’t hide that phrase after a long setup about systems theory. Search engines and database results pages are unforgiving; readers skim title lists quickly.

Clarity wins more often than ornament. Nature Index’s guidance on research paper titles quotes Elisa De Ranieri, editor-in-chief at Nature Communications, advising authors to avoid unnecessary jargon and communicate the main result clearly, ideally even to researchers new to the field (Nature Index on writing a good research paper title). Good advice. Brutally hard to follow when you’ve been living inside the project for eight months.

Length matters, but there’s no magic number. Many student guides recommend keeping titles short enough to scan while specific enough to index. In practice, 8–15 words is a useful working range. Some fields tolerate longer subtitles, especially when method and population need to be stated.

A weak title usually fails in one of four ways:

If you’re working on the whole paper, title choice should connect to the abstract and introduction. A title promises the argument; the abstract cashes the check. See research paper abstract examples and research paper introduction examples if those pieces are still rough.

Research title formulas by structure

Use formulas as scaffolding, then remove the scaffolding. If the final title still feels fill-in-the-blank, it probably needs one more pass.

Question format

Best for psychology, education, public health, and social science papers where the inquiry itself is the hook.

Formula: Does [variable] influence [outcome] in [population]?

Examples:

• Does early childhood bilingualism predict executive function in school-age children?

• Does mindfulness-based stress reduction reduce burnout in emergency nurses?

• How does peer feedback shape revision quality in first-year writing courses?

• Does campaign finance reform reduce state-level corruption?

• How do trusted messengers affect vaccine hesitancy in rural communities?

Question titles work when the answer is genuinely being tested. They get weaker when the paper already makes a clear causal claim and the title pretends to be coy.

Hypothesis-driven or finding-first

Best for STEM, medicine, and quantitative social science papers with a clear result.

Formula: [Finding] through [mechanism] in [context]

Examples:

• Social isolation mediates depression risk among remote workers

• Permafrost thaw accelerates methane release in Arctic tundra soils

• Culturally responsive teaching improves math achievement in underserved schools

• Federated learning preserves privacy in distributed medical imaging

• Dopamine prediction errors shape reward learning in adolescents

This is the highest-risk format because it can overclaim. Use it only when your evidence supports the verb. “Predicts,” “mediates,” “reduces,” and “increases” are not decorative.

Comparative format

Best when the paper tests two methods, populations, theories, interventions, or models.

Formula: [Approach A] vs. [Approach B] in [context]: [outcome]

Examples:

• Machine learning vs. logistic regression in predicting hospital readmission

• Stakeholder capitalism vs. shareholder primacy: Long-term financial outcomes

• Remote and in-person peer review in undergraduate writing courses

• Phage therapy vs. antibiotics in cystic fibrosis biofilms

• Originalist and purposivist approaches to platform regulation

Comparative titles are clean because they tell the reader exactly what is being weighed. The danger is false symmetry. If your paper barely discusses Approach B, don’t put it in the title.

Problem-solution format

Best for applied fields: engineering, public health, business, education, law, and policy.

Formula: Addressing [problem] through [method/intervention] in [field/context]

Examples:

• Addressing antibiotic resistance through phage therapy in cystic fibrosis biofilms

• Reducing urban flood risk through nature-based drainage systems

• Detecting algorithmic bias in criminal sentencing risk tools

• Improving maternal mortality surveillance through linked administrative data

• Reducing supply chain labor violations through mandatory human rights due diligence

This format works when the paper is intervention-heavy. It’s less useful for purely interpretive work.

Descriptive or exploratory format

Best for qualitative research, history, anthropology, case studies, and early-stage inquiry.

Formula: [Phenomenon] in [population/context]: A [method] study

Examples:

• Workplace mental health stigma in tech startups: A qualitative interview study

• Ritual practice and community resilience in post-conflict societies

• Language revitalization among Indigenous communities: Barriers and success factors

• Digital pilgrimage and contemporary religious practice: An ethnographic study

• Corporate accountability in supply chain labor violations: A comparative legal analysis

For qualitative papers, the title should name the phenomenon before the method unless the method is the contribution. If you’re still forming the qualitative angle, this guide to qualitative research questions is a useful companion.

Mechanistic format

Best for biology, neuroscience, medicine, chemistry, and theory-heavy social science.

Formula: The role of [mechanism] in [process]: Evidence from [method/population]

Examples:

• The role of gut microbiota in anxiety: Evidence from germ-free mouse models

• The role of NLRP3 inflammasome activation in Alzheimer’s inflammation

• The role of social trust in vaccine uptake: Evidence from rural clinics

• The role of oral tradition in Homeric epithets

• The role of platform incentives in misinformation spread

Mechanistic titles earn attention when the mechanism is specific. “The role of culture in society” is too soft. “The role of kinship obligation in post-disaster housing decisions” gives the reader something to evaluate.

STEM & life sciences titles (15 examples)

STEM titles tend to reward precision. The title often names the organism, material, model, disease, instrument, or benchmark because readers use those terms to decide whether the paper is relevant.

Biology and ecology

• Habitat fragmentation reduces pollinator diversity in temperate grasslands: A meta-analysis of 47 studies

• CRISPR-Cas9 off-target effects in primary human T cells: Quantification and mitigation strategies

• Microbial succession after wildfire in boreal forest soils

• Coral bleaching thresholds under repeated marine heatwave exposure

• Invasive grass expansion alters nitrogen cycling in semi-arid ecosystems

Chemistry

• Graphene oxide nanoparticles as pH-responsive drug carriers: Synthesis and in vitro efficacy

• Catalytic mechanisms of platinum-palladium alloys in carbon dioxide reduction

• Solvent effects on perovskite film stability under humid conditions

• Electrochemical detection of lead ions using functionalized carbon electrodes

• Polymer degradation pathways in biodegradable food packaging

Physics and engineering

• Quantum entanglement in topological superconductors: Verification using Josephson junctions

• Directional recoil imaging for dark matter detection

• Machine learning-optimized design of lightweight composite aerospace structures

• Real-time structural health monitoring using edge sensors

• Thermal fatigue prediction in additive-manufactured turbine components

Medicine and computer science

• Efficacy of personalized mRNA neoantigen vaccines in advanced melanoma: A phase II trial

• Long COVID neuroinflammation: Biomarkers and therapeutic targets

• Transformer-based protein structure prediction on CASP15 benchmarks

• Federated learning for privacy-preserving medical imaging classification

• Explainable risk prediction for 30-day hospital readmission

Medical titles often include design labels because the design changes the weight of the claim. A randomized trial, retrospective cohort study, systematic review, and case series shouldn’t sound interchangeable.

Social sciences & psychology titles (12 examples)

Social science titles usually need the relationship between variables to be legible. The National Library of Medicine’s explainer on independent and dependent variables is basic but useful here: the independent variable is the presumed cause or predictor, and the dependent variable is the outcome being measured (National Library of Medicine on dependent and independent variables).

If a title doesn’t make that relationship clear, readers have to guess the study design. Bad trade.

Psychology

• Childhood trauma and adult attachment styles: A longitudinal mediation analysis

• Does mindfulness-based stress reduction reduce burnout in healthcare workers?

• Sleep quality mediates the relationship between screen time and adolescent anxiety

• Emotional regulation predicts academic persistence among first-generation college students

Sociology and economics

• Digital inequality and educational outcomes in rural communities: A structural equation model

• Intergenerational mobility in the gig economy: Evidence from platform worker surveys

• Universal basic income and labor market participation: Quasi-experimental evidence from Kenya

• Cryptocurrency volatility and portfolio diversification among retail investors

Education and political science

• Culturally responsive teaching improves math achievement in underserved middle schools

• How peer feedback shapes revision quality in collaborative writing

• Misinformation exposure predicts election polarization: A cross-national analysis

• Does campaign finance reform reduce corruption? Evidence from state-level reforms, 1990–2020

Notice how often the strongest titles name either the method or the evidence base. “Evidence from platform worker surveys” tells the reader the claim rests on survey data. “A cross-national analysis” signals scope. Small phrases carry a lot of weight.

Humanities & history titles (10 examples)

Humanities titles have more room for interpretation, but they still need discipline signals. A literary title can be elegant. It still has to name the text, author, period, archive, concept, or interpretive lens.

History

• Women in the French Resistance: Archival evidence and oral histories

• Colonial trade networks and disease transmission in eighteenth-century Atlantic ports

• Grain shortages and urban protest in late Qing China

• Black newspapers and political organizing during the Great Migration

• Mission schools and language policy in colonial East Africa

Literature and philosophy

• Postcolonial identity in contemporary African fiction: Adichie and Teju Cole

• Metafiction and narrative unreliability in David Foster Wallace’s Infinite Jest

• Epistemic justice and marginalized knowledge systems: A Fricker-inspired critique

• Does artificial intelligence possess moral agency? A phenomenological investigation

• Translation, exile, and memory in twentieth-century diasporic poetry

Art history, classics, and religious studies

• The female gaze in Renaissance portraiture: Challenging the male-centered canon

• Digital restoration and authenticity in museum conservation

• Homeric epithets as mnemonic devices: Oral tradition and manuscript variation

• Gender and power in Sophocles’ Antigone: A feminist reinterpretation

• Syncretism in Caribbean Vodou: African diaspora and Catholic ritual fusion

The humanities tolerate colons because subtitles can pair a memorable phrase with a searchable one. Use that gift carefully. One colon is enough.

Professional & applied fields titles (8 examples)

Applied fields care about consequences. The title should make the practical stakes visible without turning into marketing copy.

Law

• Algorithmic bias in criminal sentencing: Constitutional implications and remedies

• Corporate accountability in supply chain labor violations: International legal frameworks

Legal research titles often pair a problem with a doctrinal frame. If you’re working in this area, compare these with legal research examples or the steps of legal research before finalizing the title.

Business and management

• Remote work and organizational culture: How distributed teams maintain cohesion

• Stakeholder capitalism vs. shareholder primacy: Long-term financial outcomes

Business titles need sharper nouns than “performance” and “success.” Name the measure if you can: retention, margin, churn, productivity, decision quality, error rate.

Public health and environmental science

• Social determinants of maternal mortality disparities: A systems-level analysis

• Vaccine hesitancy in rural communities: Trusted messengers and intervention strategies

• Microplastic accumulation in marine food webs: Human health implications

• Nature-based solutions for urban flooding: Cost-benefit analysis across climate scenarios

Public health titles often benefit from naming the population. “Vaccine hesitancy” is broad. “Vaccine hesitancy among rural parents of children under five” is sharper, if that’s your sample.

AI-powered title generation tips

AI is good at generating options. It’s weaker at knowing which option your field will respect.

That matches broader evidence about generative AI at work. The Harvard Business School paper on the “jagged technological frontier” found that some knowledge-worker tasks are unexpectedly easy for AI while others that look simple are hard, meaning users have to learn where the tool helps through practice (HBS field experiment on AI productivity and quality). Title drafting sits near the useful side of that frontier. Final title judgment does not.

Use AI for breadth first. Paste your abstract, research question, or thesis into the model and ask for variants that emphasize different angles: mechanism, population, method, finding, policy relevance, or theoretical contribution. Don’t ask for “better titles.” Too vague. The model will decorate.

Then compare styles. Otio’s multi-window split view and per-chat model selection lets you put Claude, GPT, Gemini, or another model side by side against the same source material. That matters because one model may produce cleaner academic phrasing while another is better at search-friendly keywords.

Ask for field calibration. A useful instruction is to have the model compare your candidate title against titles from papers in your library, then explain which terms look field-standard and which sound imported from another discipline. If you’re reading papers anyway, this beats prompting from memory.

AI title generation research is moving quickly. An arXiv paper on automatic research title generation describes using language models and datasets such as CSPubSum, LREC-COLING-2024, and a Springer social-science dataset to generate titles, reporting that model outputs were generally appropriate and reliable in the tested setting (arXiv paper on automatic generation of research paper titles). Useful. Still not a license to publish the first output.

A practical workflow:

1. Draft one plain title yourself, even if it’s ugly.

2. Generate 10 variants across the six structures above.

3. Keep the three that match your actual claim.

4. Search each candidate in Google Scholar, PubMed, JSTOR, SSRN, or your target journal site.

5. Cut hedging, filler, and unsupported novelty claims.

6. Send the final two to your advisor or co-authors.

You’ll notice a pattern after a few rounds: AI often adds “exploring,” “investigating,” “toward,” and “a comprehensive analysis of.” Sometimes those words belong. Usually they’re packing peanuts.

If you use AI for research, keep the title tied to the paper’s evidence. The title is a claim-management problem, not a branding exercise.

Common title pitfalls to avoid

The most common title mistake is vague scope. “A study of climate change” could be a middle-school essay, a Nature paper, or a policy memo. “Permafrost thaw accelerates methane release in Arctic tundra: A 20-year time-series analysis” tells the reader what changed, where, and how the evidence was gathered.

Another mistake: burying the lead behind the method. “A qualitative interview study of workplace mental health stigma in tech startups” is acceptable, but “Workplace mental health stigma in tech startups: A qualitative interview study” reads better because the phenomenon comes first.

Overstuffed colon titles are easy to spot. “Title: Subtitle: Sub-subtitle” looks like three committees negotiated a ceasefire. Use one subtitle if it adds method, context, or evidence.

Jargon without context kills reach. “Epigenetic dysregulation in NLRP3 inflammasome activation” may be fine for a narrow immunology audience. If the paper needs broader medical readers, add the disease or process: “Epigenetic regulation of NLRP3 inflammasome activation in Alzheimer’s disease.”

Be careful with novelty language. “First study,” “new framework,” and “novel approach” invite a reviewer to prove you wrong. If the contribution is real, the concrete method or dataset can carry it.

Acronyms need judgment. SARS-CoV-2 is safe in most biomedical contexts. A local program name, policy acronym, or instrument abbreviation usually needs expansion unless every likely reader knows it.

This also breaks when collaborators disagree on audience. One co-author wants PubMed discoverability. Another wants a title that reads well in a conference program. Decide the venue first, then tune the title.

How to use this list and next steps

Start by choosing the title structure that matches your evidence. If you tested a relationship, use a question or finding-first title. If you compared two models, don’t hide the comparison. If you interpreted a text, archive, doctrine, or ritual practice, name the object and lens.

Next, draft three to five candidates. One should be painfully plain. One should be more field-specific. One can use a subtitle. The plain version often reveals what the fancy version is trying to avoid.

Then validate the keywords. Search the exact title and the two or three core phrases. If your title uses terms nobody in the field uses, fix that before you polish the wording. This is where how to search for research articles becomes part of title writing, not a separate task.

For longer papers, check that the title, thesis, abstract, and conclusion agree with each other. A title that promises causal evidence shouldn’t lead to a descriptive essay. If you need to tighten the rest of the structure, use the research paper structure guide.

A final pass should be boring in the best way. Remove filler. Replace vague nouns. Keep the method only if it helps the reader understand the claim.

Try Otio for your next research title draft when you want title variants grounded in your own sources.

FAQ

Q: How long should a research title be?
A: Aim for 8–15 words as a working range. Some journals and disciplines allow longer subtitles, but long titles should earn every word.

Q: Should my title be a question or a statement?
A: Both can work. Questions fit psychology, education, and exploratory social science; declarative titles are common in STEM and medicine when the evidence supports a clear finding.

Q: Can I use AI to generate my entire title?
A: Use AI to generate options, then choose and revise based on your actual evidence, target journal, and field conventions. Don’t let the model add novelty or causal claims your paper doesn’t prove.

Q: How do I avoid plagiarizing another paper’s title?
A: Search the final title in Google Scholar, PubMed, JSTOR, SSRN, or the target journal’s site. If an identical or very close title exists, adjust your population, method, evidence, or angle.

Q: What if my title is too long for the journal’s limit?
A: Cut filler first: “novel,” “comprehensive,” “preliminary,” and “an investigation of” often disappear without loss. If needed, move secondary details into the abstract.

Related reading

• How To Title A Research Paper in 6 Steps

• 100 Research Questions Examples For Students

• 121 Examples Of Research Topics for Inspiration

• 3 Research Paper Abstract Examples + Writing Tips

• Research Paper Introduction Examples, Format, and Tips

• Complete Guide On Research Paper Structure

• Full Guide to Using AI for Research: Tips & Tools

• 28 Most Useful Tips For Writing Research Papers

• What Is A Qualitative Research Question

• 6 Effective Tips for Searching Research Articles