Ethical considerations in research: Best practices and examples

Ethical considerations in research form the backbone of credible and responsible scientific inquiry. They guide how you design studies, collect and handle data, interact with participants, and share your results. These principles keep participants safe, make your data quality stronger, and build trust in your findings.

This guide covers the key ethical principles that shape responsible research. We’ll walk you through how they apply across every stage of a project and how they’re evolving in response to new challenges, like AI-generated content. 

You’ll discover: 

• Examples of ethical considerations in research
• Best practices for informed consent and privacy
• Strategies to carry out research in line with both regulatory and ethical requirements

What are ethical considerations in research?

Research ethics are the principles and standards that govern how you run scientific studies. They:

• Protect participants by safeguarding their rights, privacy, and well-being. 
• Protect you and the institutions you work for with clear standards that ensure your findings are valid. 

Today’s research ethics principles are built on three core values established in the Belmont Report: 

1. Respect for persons: Treating people with autonomy. They can make informed decisions about whether they want to take part. 
2. Beneficence: Maximizing benefits while minimizing potential harm. 
3. Justice: Ensuring that the burdens and benefits of research are spread fairly across populations.

Research integrity depends on ethical practices at every stage, from initial study design through publication and data sharing. 

The six core research ethics principles

Ethical research is based on six key principles. Each one helps you navigate the responsibilities and risks of working with human participants. 

1. Autonomy and informed consent: Participants should be able to make voluntary, informed decisions about joining a study. That means giving them clear details about what will happen, any risks involved, and any potential benefits, without pressure or coercion.
2. Beneficence: Beneficence in research is about making sure your work benefits the world while minimizing harm. You’ll need to think carefully about possible risks and put strong safeguards in place to protect participants from harm.
3. Integrity and scientific validity: This means being honest and transparent in how you conduct your research. Collect data accurately, report results truthfully, give proper credit, and avoid any form of fabrication, falsification, or bias.
4. Justice: Calls for fairness in how research burdens and benefits are shared. Vulnerable groups shouldn’t shoulder more risk than others, and the benefits of research should reach the communities who need them.
5. Confidentiality and data protection: Participants trust researchers with personal information, and that trust needs to be protected. Secure data storage, clear access controls, and strong retention policies help ensure their privacy is respected. 
6. Accountability and oversight: Institutional review boards (IRBs) or research ethics committees (RECs) make sure studies meet ethical standards. These groups review study plans before work begins and continue monitoring compliance as research progresses.

This ethical framework keeps participants safe and makes the quality and credibility of your research stronger. 

Ethical considerations in practice

You need to apply ethical principles across the entire research process, from initial planning through data analysis and publication. 

Voluntary participation and informed consent

Participants should join a study because they genuinely want to, not because they feel pressured or obligated. They also need to be able to step away at any point without facing any consequences. This freedom is what distinguishes ethical research from coercive or exploitative practices.

Informed consent in research goes hand in hand with voluntary participation. Before a study begins, you need to provide clear, complete information so participants understand exactly what they’re agreeing to. A strong consent process includes:

• A clear explanation of the study’s purpose: Describe your goals and methods in everyday language.
• An honest description of potential risks and benefits: Even small risks should be acknowledged.
• Accurate expectations about time commitment: Let participants know how long the study will take and flag any multiple sessions.
• Accessible contact information: Share details for the researcher, the institution, and any study sponsors so participants can easily ask questions.
• A reminder of their right to withdraw: Make it clear that they can leave the study at any time without affecting their relationship with you or the institution.
   

Ethics committees (IRBs or RECs) review these consent procedures to ensure they’re fair, transparent, and protective of vulnerable groups. They look at whether the language is easy to understand and whether researchers have a plan for renewing consent in longer-term studies.

Cultural sensitivity in research

Cultural sensitivity helps to ensure your research is inclusive, valid, and respectful. It also strengthens the quality of your results. Approaches that work well in one cultural setting may not translate smoothly to another. So, it’s important to take differences seriously and plan with care.

Here are a few examples of inclusive research practices:

• Schedule sessions with cultural and religious observances in mind
• Offer study materials in participants’ native languages
• Be mindful of dietary restrictions, communication norms, and social hierarchies that might influence how participants interact with your research

Culturally diverse research teams are another key factor here. When team members come from a variety of backgrounds, they can help find cultural blind spots, suggest ways to adapt, and help interpret findings within the right cultural context.

Privacy, anonymity, and data protection

Data protection in research involves both strong technical safeguards and thoughtful study design. A big part of this is understanding the difference between anonymity and confidentiality, and choosing the right approach for your project.

Anonymity vs. confidentiality

Anonymity means removing any data that could identify a participant. This includes:

• Names
• Addresses
• Email addresses
• Photos
• Video footage

In an anonymous dataset, identifiers are replaced with codes that can’t be traced back to individuals. If the data is truly anonymous, even the research team can’t link responses to specific participants.

Confidentiality, on the other hand, is about protecting identifiable data. This requires secure systems and limited access. Good practices include:

• Storing data on encrypted servers with strong authentication and regular security updates
• Keeping participant identifiers in a separate database, connected to the data only through coded references

GDPR research compliance adds another layer of responsibility. You’ll need clear data management policies that explain how long data will be stored and when it will be deleted. Access should be restricted to essential team members, and data processing agreements should be in place when working with external partners.

Pseudonymization offers a useful middle ground. Here, direct identifiers are swapped for pseudonyms, but a separate key is kept so participants can be re-identified if needed - helpful for longitudinal studies and still more protective than fully identifiable data.

Finally, be mindful of how you share your findings. Avoid discussing specific cases in public settings where details might unintentionally reveal someone’s identity. Focus on aggregated results and general patterns to keep participants’ privacy intact.

Authenticity and research integrity

Research integrity is about being honest at every stage of the process, from collecting data to publishing your results. This means reporting findings accurately, being upfront about limitations, and steering clear of plagiarism, fabrication, or falsification.

The growing use of AI tools like ChatGPT adds a new layer of complexity. Researchers now need to tell the difference between genuine human responses and those generated or copied with the help of AI. 

Tools like Prolific’s authenticity checks help with this by using behavioral patterns to identify AI-generated content. These checks catch AI responses with 98.7% accuracy, while keeping false positives extremely low at 0.6%.

You can reinforce this by giving clear, explicit instructions reminding participants not to use AI tools or external sources. Doing so can reduce AI misuse by up to 61%, and it helps set expectations that support fairness and transparency.

Integrity is also a part of how you document and share your work. Properly citing sources, describing your methods clearly, and sharing data when appropriate all contribute to a more trustworthy research process. Practices like preregistering studies, embracing open science, and supporting replication efforts make research more transparent and help prevent selective reporting or cherry-picking results.

Managing conflicts of interest

Conflicts of interest in research can crop up when personal, financial, or professional relationships influence how studies are run or interpreted. Having a conflict doesn’t automatically undermine a study. But failing to disclose it can hurt credibility and trust.

To be transparent, make sure you clearly state any potential conflicts in your research proposals, publications, and presentations. These might include:

• Financial ties to companies that could benefit from certain outcomes
• Personal relationships with participants or institutions connected to the study
• Professional affiliations that might create pressure to reach specific conclusions

There are also practical steps you can take to reduce the impact of conflicts on your work:

• Bring in independent data analysts who have no stake in the findings.
• Use blinding so researchers assessing outcomes don’t know which participants received which conditions.
• Create your analysis plan before you look at the data to avoid being influenced by early results.

Independent oversight adds another layer of protection. Ethics committees, peer reviewers, and advisory boards can help assess whether conflicts pose a risk to objectivity. When needed, they’ll recommend extra safeguards or advise that a researcher step back from certain parts of the project.

Minimizing potential harm

Potential harm in research doesn’t just mean physical injury. A thorough risk assessment should look at every type of harm participants might face. This includes psychological, social, physical, and legal risks. Understanding these categories helps you design studies that protect people from unintended consequences.

Psychological harm

Psychological harm can involve stress, anxiety, emotional distress, or trauma, especially in studies that explore sensitive topics such as discrimination, abuse, or loss. To help reduce psychological risk:

• Provide mental health resources
• Offer breaks during difficult or emotional procedures
• Include debriefing sessions that help participants process their experience

Social harm

Social harm affects participants’ relationships, reputation, or standing in their communities. Studies involving stigmatized identities or behaviors can put people at risk if it becomes known that they took part. To reduce social harm:

• Put strong confidentiality protections in place
• Think carefully about how you describe your findings to avoid identifying participants

Physical harm

While most common in medical or biological research, physical risks can appear in any study involving substances, equipment, or strenuous tasks. To minimize the risk of physical harm:

• Screen participants for any conditions that might increase their risk
• Monitor for adverse reactions throughout the study and be ready to respond

Legal harm

Legal harm can occur when research touches on illegal activities or produces records that might put participants at risk. To protect participants:

• Use Certificates of Confidentiality where applicable
• Handle sensitive data with extra care
• Clearly explain any legal limits to confidentiality before the study begins

Ethics committees will look at whether the benefits of your research justify the risks and whether your safeguards are strong enough. Under the beneficence principle, the benefits of your research should outweigh the potential harm.

If harm does occur despite your precautions, you’re responsible for offering appropriate support and reporting any serious adverse events in line with regulatory requirements.

To protect participant well-being, Prolific encourages researchers to include content warnings that identify where studies feature sensitive or disturbing content. You can find out more about how to use content warnings in our guide.

Ethical use of technology in research

As more research moves online and relies on digital tools and AI, new ethical questions arise around privacy, consent, and potential bias. It’s important to think carefully about how these technologies shape participant rights and the integrity of your data.

Data protection

Digital environments introduce added complexity when it comes to keeping data secure. Cloud storage, survey platforms, and mobile apps all come with their own vulnerabilities.

To help reduce these risks:

• Use end-to-end encryption for data transmission
• Require secure authentication methods
• Check that your technology vendors follow strong security standards
• Run regular security audits to identify and address emerging threats

Informed consent

Technology also changes the way participants understand and agree to take part in a study. Digital consent forms should clearly explain how tools will be used, where data will be stored, and who can access it. They should offer the same level of detail as traditional consent forms, plus easy ways for participants to ask questions or withdraw electronically.

Algorithmic bias

AI tools can introduce bias, especially when they’re trained on data that doesn’t represent all groups fairly. This can lead to inaccurate or unequal results.

If you’re using AI in your research, make sure to:

• Evaluate algorithms for potential bias
• Test how well they perform across different demographic groups
• Be transparent about any limitations so readers can interpret your findings responsibly

Fair treatment and compensation

Fair treatment starts with respecting participants’ time and contributions. That includes compensating them appropriately. While payment is part of the picture, fairness also involves being transparent, accessible, and respectful throughout the entire research experience.

Compensation should reflect the effort, time, and expertise participants bring to your study. Prolific sets a minimum rate of $8.00 / £6.00 per hour, though we recommend $12.00 / £9.00 per hour as a more appropriate baseline. It’s also important to check your institution’s guidelines, since some have their own rules around minimum or maximum rates.

Here are a few factors that can help guide fair compensation in research:

• Task complexity: Recording videos, completing cognitive tests, or providing expert feedback takes more effort than simple surveys. Pay should match the level of demand you’re placing on participants.
• Who you’re recruiting: If you need people with specific skills or backgrounds, you may need to offer higher participant incentives to attract a big enough sample, while still avoiding payments so high that they feel coercive.
• Transparency around incentives: Clearly communicate payment amounts, timelines, and methods in both recruitment materials and consent forms. Let participants know if anything (like incomplete responses or missed attention checks) could affect payment, so expectations are clear from the start.

Oversight and accountability

Most research that involves human participants needs approval from an independent ethics review board. This external review helps ensure your study meets ethical standards and genuinely protects the people taking part. In the United States, these reviews are handled by Institutional Review Boards (IRBs); in many other countries, they’re carried out by Research Ethics Committees (RECs).

Before you begin, the committee will look closely at your study proposal. They’ll weigh potential risks and benefits, assess your consent procedures, and check that you have reliable plans in place to protect participants’ privacy and wellbeing.

Here’s a quick look at how the review process typically works:

Submission

You’ll start by submitting a detailed protocol that outlines your research objectives, methods, recruitment approach, consent process, and data protection measures. After reviewing it, the committee might approve the study, request revisions, or, in rare cases, determine that the risks are too great.

Risk assessment

Risk assessment is central to the review:

• Minimal risk studies (where the risks are no greater than those of everyday life) may qualify for an expedited review.
• Studies with greater-than-minimal risk usually require a full committee evaluation.
• Some research may be exempt if it poses little to no risk and doesn’t involve vulnerable populations.

Ongoing monitoring

Ethical oversight doesn’t end with approval. Throughout your study, you’ll need to report any changes to the protocol, note any adverse events, and submit updates when the study ends. Many committees also conduct annual reviews to confirm that everything remains compliant and that no new risks have emerged.

Ethics oversight is there to support both you and your participants. Approval from an independent committee gives you clear documentation that ethical standards were met (helpful if questions arise later), and it gives you the chance to address potential issues before they cause harm.

For more information about IRB/REC applications and ethics approval, read our guide.

Common ethical challenges in modern research

As new technologies reshape how we run research, they also raise questions that traditional guidelines don't always have an answer to. Online and digital environments, in particular, come with some ethical issues that you’ll need to approach with care.

Here are some of the most common modern ethical challenges that you’ll face, and some tips on how to overcome them. 

AI tools and participant deception

Participants may turn to tools like ChatGPT to help generate their responses, which can compromise the authenticity of the data.

Solution: Use behavioral checks to spot AI-generated content, and be clear upfront that participants shouldn’t use AI assistance.

Social media data scraping

Even though social media posts are public, many users still think of them as private or personal spaces. They may not expect their content to be collected for research.

Solution: Think carefully about whether informed consent is feasible and whether your intended data use fits both platform rules and user expectations.

Cross-border data sharing and GDPR

International research brings together teams working under different data-protection laws. This can create practical and legal complications.

Solution: When in doubt, follow the most protective privacy standard across all participating regions, and get explicit consent for any cross-border data transfers.

Online consent and participant anonymity

Digital consent processes can feel impersonal. Participants may skim through them without fully understanding what they’re agreeing to.

Solution: Add simple comprehension checks. Make contact details easy to find and give participants space to ask questions before they consent.

Algorithmic bias in automated systems

Automated tools used for recruitment or data analysis can unintentionally disadvantage certain groups if they’re not designed or tested carefully.

Solution: Regularly test these systems for demographic bias. Back up automated decisions with human oversight.

Participant fatigue in long-term online studies

Running studies remotely makes frequent data collection easier. But that convenience can sometimes place too much burden on participants over time.

Solution: Keep data requests to what’s genuinely necessary. Make sure ongoing participation is fairly compensated.

Examples of ethical and unethical research

We’ve explored some of the most important ethical principles in research. But what do they look like in practice? What happens when researchers do - or don’t - follow them? 

Let’s examine a real-world example of ethical research, plus some unethical research examples that show why it’s so important to follow these principles. 

Ethical example: COVID-19 vaccine trials

During the pandemic, major pharmaceutical companies moved quickly, but not carelessly, to run vaccine trials. They put a lot of care into informed consent, giving participants clear, detailed explanations about the experimental vaccines, possible risks, and their absolute right to withdraw at any time.

Independent safety boards were reviewing data constantly, and the trials worked hard to include diverse participants so results would reflect the wider population. Throughout the process, researchers reported both good and bad findings openly, which was essential for maintaining public trust at such a high-stakes moment. 

Unethical example: Tuskegee Syphilis Study

From 1932 to 1972, the US Public Health Service carried out a study in which treatment was deliberately withheld from Black men with syphilis. Participants weren’t told they had the disease, nor were they informed when penicillin became the standard treatment.

This study caused immense harm and represented a profound breach of the most basic ethical principles. Its legacy has shaped many of the modern protections and regulations we rely on today.

Unethical example: Facebook emotional contagion experiment 

In 2014, Facebook altered the news feeds of nearly 700,000 users to test whether emotions could spread online. Some people were shown more positive posts, others more negative ones. But none were told an experiment was happening, and the only “consent” was buried in the general terms of service.

The study sparked widespread concern about what it means to experiment on people through digital platforms, especially when they don’t even know they’re part of a study.

FAQs

What are the main ethical principles in research?

The core ethical principles in research include: 

• Respect for persons (autonomy and informed consent)
• Beneficence (maximizing benefits while minimizing harm)
• Justice (fair distribution of research burdens and benefits)
• Integrity (honest and valid research practices)
• Confidentiality (protecting participant privacy)
• Accountability (independent ethics review) 

These principles work together to protect participants and maintain the credibility of the research.

What is informed consent?

Informed consent means participants voluntarily agree to participate after receiving comprehensive information about the study. A proper consent process includes: 

• Clear explanations of research purpose and methods
• Potential risks and benefits
• Time commitment
• Contact information for researchers
• The right to withdraw at any time without penalty

Consent must be given freely without coercion or undue inducement.

When do I need IRB approval?

IRB or REC approval is typically required for research involving human participants at institutions that receive government funding or are conducted by researchers at universities and hospitals. The need for approval depends on factors including participant risk level, study setting, and institutional policies. 

Even minimal-risk research usually requires at least an expedited review. Check with your institution's research office to determine requirements for your specific study.

How can researchers protect confidentiality?

Researchers protect confidentiality through multiple safeguards, including: 

• Encrypted data storage
• Access controls that limit who can view participant information
• Separation of identifying information from research data
• Use of coded identifiers
• Clear data retention policies
• Secure data destruction once information is no longer needed

GDPR compliance requires explicit documentation of data handling procedures and participant consent for data use.

What are common ethical issues in online studies?

Digital research presents ethical challenges in remote settings, including: 

• Informed consent
• Privacy protection for data collected through apps and platforms
• Algorithmic bias in automated systems
• The authenticity of responses (particularly with AI tools)
• Cross-border data transfers that must comply with multiple regulations
• Fair compensation for online participants 

There are various methods you can use to tackle these challenges, including behavioral authenticity checks, explicit consent procedures, and robust data security measures.

Why ethical research matters

Ethical research practices help protect participants from harm, support research credibility, and sustain the public trust that research depends on. When researchers ground their work in strong ethical principles, their results carry more weight and can be used with confidence to guide important decisions in society.

Ethics also play a practical role: it shields researchers and institutions from legal risk and encourages responsible progress that benefits everyone. Rather than being separate from the research process, ethics is woven into every part of it, shaping work that is both responsible and meaningful.

Prolific supports ethical research practices with: 

• An industry-leading reward structure for fair compensation 
• Authenticity checks that ensure data quality without unfairly penalizing participants
• Tools designed around participant welfare

Find out more about how we ensure high-quality, ethical human data for your research.