How to Design a Survey for a Research Study

TL;DR: This post answers a specific and practical question: how do you design a survey that produces data you can actually analyse and publish? It is written for high school students in grades 9 through 12 who are conducting original social science, psychology, public health, or interdisciplinary research. By the end, you will know how to write valid survey questions, structure your instrument, avoid the most common design errors, and prepare your data for submission to a peer-reviewed journal. If your research is ready for review, the Princeton Journal of Pre-Collegiate Research publishes original student work across all disciplines, including survey-based studies.

Why survey design is where most student research fails

Survey-based research is one of the most accessible methods available to high school students, but it is also one of the most frequently mishandled. The most common reason peer reviewers reject survey studies is not a weak research question. It is a flawed instrument: questions that are double-barrelled, scales that are inconsistent, or samples so small they cannot support the conclusions drawn. Knowing how to design a survey for a research study is not a minor technical detail. It is the foundation on which your entire argument rests. If the instrument is compromised, the data is compromised, and no amount of careful analysis can recover it.

How do you design a survey for a research study?

To design a survey for a research study, you must complete five steps in order: define your research question precisely, operationalise your variables, write individual items using validated formats, structure the instrument logically, and pilot-test it before collecting real data. Each step builds on the previous one. Skipping any of them produces a survey that generates noise, not evidence.

Step 1: Define your research question before writing a single question

Your survey exists to answer one specific research question. Before you write any items, write that question down in one sentence. If you cannot state it in one sentence, your study is not yet focused enough to design an instrument.

A strong survey research question specifies the population, the variable being measured, and the relationship being tested. For example: "Does reported screen time before bed correlate with self-reported sleep quality among high school students aged 14 to 18?" That sentence tells you exactly who your respondents are, what you are measuring (screen time and sleep quality), and what relationship you are investigating.

A weak question like "How does technology affect teenagers?" cannot be operationalised into a survey. It is too broad to produce measurable data. Narrow it before you proceed.

Step 2: Operationalise your variables

Operationalisation means defining exactly how you will measure each variable in your study. This is where most student surveys collapse. A concept like "stress" or "academic motivation" cannot be measured directly. You need to specify the observable indicators you will use as proxies.

The most reliable approach at the high school level is to adapt a validated scale that researchers have already tested for reliability. The Perceived Stress Scale (PSS-10), developed by Cohen, Kamarck, and Mermelstein (1983), is a widely used and freely available instrument that measures perceived stress across ten items. Using an established scale, and citing it properly, signals methodological credibility to reviewers. You can learn more about how to present your methodology clearly by reading this guide on how to write a research methods section.

If no validated scale exists for your specific variable, you will need to construct your own items and justify your choices in the methods section. That justification must explain why each item maps onto the underlying construct you are trying to measure.

Step 3: Write your survey items using validated formats

Individual survey items must follow specific rules to produce usable data. These are not stylistic preferences. They are methodological requirements.

Use a consistent Likert scale for attitudinal or frequency questions. A five-point or seven-point scale (for example: 1 = Strongly Disagree to 5 = Strongly Agree) is standard in social science research. Do not mix a five-point scale in one section with a ten-point scale in another. Inconsistency makes your data impossible to compare across items.

Write one idea per item. A question like "Do you feel stressed and tired after school?" is double-barrelled: a respondent who feels stressed but not tired cannot answer accurately. Split it into two separate items.

Avoid leading language. "Do you agree that social media is harmful to mental health?" pushes respondents toward a particular answer. A neutral framing would be: "How often do you feel that your social media use negatively affects your mood?"

For demographic questions, use defined categories rather than open-ended fields where possible. "What is your age?" produces cleaner data if you provide ranges (14-15, 16-17, 18-19) rather than a blank text box.

Step 4: Structure the instrument logically

The order of your questions affects how respondents answer them. This is called order bias, and it is well-documented in survey methodology literature. Place demographic questions at the end, not the beginning. Starting with age, gender, or school year can prime respondents to think about their identity before answering substantive questions, which can skew results.

Group thematically related items together. If your survey covers sleep habits, screen time, and academic performance, keep each theme in its own section with a clear header. Jumping between topics increases cognitive load and reduces response quality.

Include an informed consent statement at the start of the survey. For student research, this is not optional. It must explain the purpose of the study, confirm that participation is voluntary, and state how data will be stored and used. If your respondents are minors, parental consent may also be required depending on your school or institution's ethics guidelines.

Step 5: Pilot-test before you collect real data

A pilot test means administering your survey to a small group (5 to 10 people) who are similar to your target population, but who are not part of your actual sample. Ask them to complete the survey and then tell you which questions were confusing, ambiguous, or difficult to answer.

Pilot testing catches problems that you cannot see yourself because you already know what each question means. A respondent who does not share your assumptions will find the gaps. Fix every item that produces confusion before you launch the full study.

What makes a survey-based study publishable rather than just descriptive?

Collecting survey data is not the same as conducting research. The difference lies in what you do with the data and how you frame the contribution. Reviewers at peer-reviewed journals assess whether your study answers a question that was genuinely unanswered before, or tests a hypothesis with enough methodological rigour to support a conclusion.

A descriptive survey that simply reports percentages ("62% of respondents said they felt stressed") does not constitute original research unless it is measuring a population that has not been studied before, or using a comparison that produces new insight. What elevates a survey study is analysis: correlation, regression, comparative analysis between subgroups, or longitudinal tracking. The analysis must match the research question stated in your introduction.

Sample size also matters. A survey of 15 classmates cannot support generalised conclusions about high school students. While large nationally representative samples are not realistic for most student researchers, a sample of 80 to 150 respondents drawn from a defined and described population is a defensible starting point for a student study, provided you acknowledge the limitations explicitly. For guidance on how to frame your argument around the data you have, the post on data vs. evidence: what reviewers look for in student research is worth reading before you begin your analysis.

What are the most common survey design mistakes high school students make?

The most common survey design mistakes are: using unvalidated scales without justification, writing double-barrelled or leading questions, collecting too small a sample to support the stated conclusions, and skipping the pilot test entirely. Each of these errors is fixable before data collection begins, but almost impossible to correct after.

The first mistake is treating survey design as straightforward. Students often write questions based on intuition rather than methodology. The result is an instrument with low construct validity: it measures something, but not necessarily what the researcher intended to measure. The fix is to ground every item in a defined construct and cite the methodological source for your scale choices.

The second mistake is convenience sampling without acknowledging its limits. Surveying only your own school, friend group, or social media followers produces a biased sample. This is not automatically disqualifying, but you must name the limitation in your discussion section and explain why the findings may not generalise. Reviewers will raise this if you do not address it first. See the guide on what makes a research paper get rejected for a fuller breakdown of how methodological gaps lead to rejection.

The third mistake is drawing causal conclusions from correlational data. A survey can establish association. It cannot establish causation. If your data shows that students who report higher screen time also report lower sleep quality, you can say these variables are correlated. You cannot say screen time causes poor sleep without experimental evidence. Overstating your findings is one of the fastest routes to a rejection decision.

How to design a survey for a research study, step by step

1. Write your research question in one sentence. Confirm it specifies a population, a variable, and a relationship.

2. List every variable your study requires. For each one, identify either an existing validated scale or define your own items with explicit justification.

3. Draft your survey items. Apply the one-idea-per-item rule. Use a consistent Likert scale. Remove all leading language.

4. Write your informed consent statement. Place it at the start of the survey. Confirm it meets your school or institution's ethics requirements.

5. Structure the instrument: substantive questions first, demographic questions last. Group items by theme.

6. Run a pilot test with 5 to 10 people. Revise every item that generates confusion or ambiguity.

7. Collect your data from a clearly defined and described sample. Record your sample size and recruitment method for the methods section.

8. Analyse your data using a method appropriate to your research question (correlation, comparative analysis, descriptive statistics with subgroup breakdowns).

9. Write up your findings with explicit acknowledgement of limitations. If your sample is small or non-representative, say so.

10. When your paper is complete, review the submission guide for high school student researchers and prepare your manuscript for peer review.

PJPCR publishes original survey-based research across social sciences, psychology, public health, and interdisciplinary fields. If your work is ready for peer review, review the submission guidelines at how to format a research paper for publication before you submit.

Frequently asked questions about how to design a survey for a research study

What is a Likert scale and why do researchers use it?

A Likert scale is a rating scale that measures the degree of agreement, frequency, or intensity of a response, typically across five or seven points. Researchers use it because it converts subjective attitudes into ordinal data that can be analysed statistically. A five-point scale ranging from "Strongly Disagree" to "Strongly Agree" is the most common format in social science research.

Likert scales are preferred over yes/no questions because they capture nuance. A respondent who somewhat agrees with a statement provides different information than one who strongly agrees. That distinction matters when you are testing a hypothesis about degree or intensity rather than binary presence or absence.

How many questions should a high school research survey have?

Most peer-reviewed survey studies at the student level use between 15 and 40 items, depending on the number of variables being measured. Shorter surveys produce higher completion rates. Longer surveys allow for more validated sub-scales but risk respondent fatigue, which degrades data quality toward the end of the instrument.

A practical target for a focused student study is 20 to 30 items. This is enough to measure two or three constructs with validated scales while keeping the completion time under 10 minutes. If your survey exceeds 40 items, review your research question. You may be trying to measure too many things in one study.

Do I need university lab access to conduct survey research in high school?

No. Survey research is one of the most accessible methodologies for high school students precisely because it does not require laboratory equipment, institutional affiliation, or specialised software. You can design, distribute, and analyse a survey using freely available tools such as Google Forms for data collection and standard spreadsheet software for basic statistical analysis.

What you do need is a clearly defined research question, a defensible sampling strategy, and methodological transparency in your write-up. The absence of a university affiliation does not disqualify a study. Reviewers assess the quality of the methodology, not the prestige of the institution behind it.

What makes a survey-based research paper publishable in a peer-reviewed journal?

A publishable survey study addresses a specific research question that is not already answered in the existing literature, uses a validated or clearly justified instrument, collects data from a defined and described sample, applies appropriate statistical analysis, and explicitly acknowledges its limitations. Descriptive results alone are rarely sufficient for publication.

The discussion section is where most student papers lose credibility. Reviewers expect you to interpret your findings in relation to prior research, not just restate what the data shows. If your results contradict an existing study, explain why. If they confirm one, explain what your study adds. A paper that situates its findings in the broader literature is substantially more likely to pass peer review.

What kinds of survey research does PJPCR publish?

The Princeton Journal of Pre-Collegiate Research publishes original survey-based research across social sciences, psychology, public health, education, environmental studies, and interdisciplinary fields. Submitted work must be original, methodologically sound, and not previously published elsewhere. Peer review is conducted by qualified reviewers, and acceptance is selective.

The standard review and publication timeline is 2 to 3 months. A fast-track option is available for students who need a quicker turnaround. Submission and peer review are free. A publication fee applies for accepted papers. To understand what reviewers assess during the process, read the guide on what happens after you submit your research paper.

Design the instrument before you collect a single response

Survey design is not a step you can revisit after data collection begins. Every decision you make before distribution, from how you frame each item to how you structure the consent statement, determines the quality of the evidence you will have to work with. Validate your constructs. Pilot-test your instrument. Define your sample before you recruit it. Acknowledge your limitations before a reviewer has to point them out.

Research that is methodologically honest is research that holds up under scrutiny. That is the standard peer review applies, and it is the standard worth meeting. If your survey-based study is complete and ready for that scrutiny, submit it to PJPCR through the submission guidelines for high school student researchers.