A survey is the most common primary research method in undergraduate marketing research assignments — and the most commonly done poorly. The problem is not that students cannot write questions. The problem is that most students write questions that feel reasonable but introduce bias, generate data that cannot be analysed academically, or cannot be justified in the methodology section because no thought was given to the research design before the survey was built.
This guide covers the full survey design process for a marketing research assignment — from deciding whether a survey is the right method for your research question, through question type selection and bias elimination, to how you actually analyse and present the data in your findings section. Each section includes worked examples you can apply directly to your own assignment.
When Should You Use a Survey in a Marketing Research Assignment
Not every marketing research question is suited to a survey. Using a survey when a different method would be more appropriate is itself a methodological weakness — and if your brief asks you to justify your research design, choosing the wrong instrument will cost marks before you have written a single question.
Use a survey when your research question requires:
- Measurable, comparable data across a group of respondents — attitude scores, frequency, preference rankings
- Pattern identification across demographic variables — do younger consumers respond differently to older ones?
- Quantification of a behaviour or opinion — what percentage of consumers would pay a premium for sustainable products?
Do not use a survey as your primary method when your research question asks "why" rather than "how many" or "how often." Why questions require qualitative methods — interviews or focus groups — because they explore meaning and motivation that a tick-box survey cannot capture.
| Research Question Type | Right Method | Wrong Method |
|---|---|---|
| "What percentage of consumers prefer Brand A over Brand B?" | Survey ✓ | Interview |
| "How frequently do students purchase fast fashion items?" | Survey ✓ | Focus group |
| "Why do consumers feel loyal to a particular brand?" | Interview ✓ | Survey |
| "How do consumers experience the online checkout process?" | Focus group ✓ | Survey |
| "Which marketing channel influences purchase intent most strongly?" | Survey ✓ | Observation |
Mixed methods note: You can combine a survey (quantitative) with a small open-ended section (qualitative) to address both "how many" and "why" within a single instrument. This is called a mixed-methods approach and is well-regarded at undergraduate level — provided you justify the combination in your methodology. The example assignment in Blog 34 uses exactly this approach.
How to Structure Your Survey — Question Types Explained
The question types you use determine what kind of data you collect — and what kind of analysis you can present in your findings section. Choosing question types without thinking about analysis first is one of the most common survey design mistakes. Each type below is explained with its academic purpose and a worked example relevant to a marketing research assignment.
Likert Scale
Measures attitude or agreement on a 5 or 7-point scale. Produces ordinal data that can be statistically analysed and presented as mean scores or frequency distributions.
Multiple Choice (MCQ)
Offers a fixed set of options. Easy to analyse — produces frequency counts and percentages. Good for demographic segmentation and categorical variables.
Open-Ended
Captures qualitative responses in the respondent's own words. Harder to analyse — requires thematic coding. Adds depth but cannot be statistically summarised.
Dichotomous (Yes/No)
Binary response. Produces clear, simple data. Useful as filter questions — screening out irrelevant respondents before more detailed questions.
How a Likert Scale Question Should Look
The Likert scale is the most academically important question type in a marketing survey because it generates the kind of measurable, comparable data that supports statistical analysis in your findings section. Here is what a correctly designed Likert scale question looks like:
Disagree
1
2
Agree nor
Disagree
3
4
Agree
5
5-point vs 7-point Likert scale: Use a 5-point scale for undergraduate assignments — it is simpler to analyse and present, and the difference in data quality is negligible at small sample sizes. 7-point scales are more appropriate at postgraduate level where statistical analysis is more sophisticated.
How Many Questions and Respondents Does Your Survey Need
Two questions every student asks before designing their survey — and two questions most students answer incorrectly by defaulting to round numbers with no justification.
Number of Questions
The right number of questions is determined by your research objectives — not by what feels substantial. A survey that is too long generates incomplete responses and respondent fatigue. A survey that is too short cannot answer the research question adequately.
- 8–12 questions is the optimal range for an undergraduate marketing research assignment survey
- Start with 2–3 demographic questions (age, gender, location or similar relevant variable)
- Include 4–6 Likert scale questions covering your key research variables
- Include 1–2 multiple choice questions for categorical data
- End with 1 open-ended question if using mixed methods
- Completion time should be under 5 minutes — state this in your survey introduction to improve response rate
Number of Respondents
For an undergraduate marketing research assignment, 30–50 respondents is the accepted minimum for quantitative survey data. Below 30, your data is too thin to identify patterns reliably. Above 50, you are investing time in data collection that could be better spent on analysis and writing.
Critical methodology point: Whatever your sample size, you must justify it in your methodology section — not just state it. Write: "A convenience sample of 45 respondents was recruited via [channel], which is appropriate for an exploratory undergraduate study where the objective is pattern identification rather than population-level inference (Bryman, 2016)." This one sentence protects your methodology from the most common marker criticism of small-sample primary research.
| Sample Size | Academic Status | What to Say in Methodology |
|---|---|---|
| Under 30 | Weak — hard to justify statistically | Frame as pilot study; acknowledge as major limitation |
| 30–50 | Acceptable at undergraduate level | Justify as exploratory; cite Bryman (2016) |
| 50–100 | Strong for undergraduate work | Justify as sufficient for pattern identification |
| 100+ | Excellent — enables basic statistical comparison | Can make limited generalisability claims with caveats |
How to Write Survey Questions That Do Not Bias Your Results
Biased survey questions are the single most common technical failure in undergraduate marketing research assignments. The problem is that biased questions often sound perfectly reasonable — the bias is subtle, built into word choice or question structure, and invisible unless you know what to look for. Markers do know what to look for.
The six most common question bias types in marketing assignment surveys — with a biased example and a corrected version for each:
1. Leading Questions
2. Double-Barrelled Questions
3. Loaded / Assumptive Questions
Pilot test rule: Before distributing your survey, send it to 3–5 people and ask them to flag any question where they felt unsure how to answer, or where they felt the question was guiding them. Their feedback will identify bias you cannot see yourself. Mention this pilot test in your methodology — it demonstrates research rigour and earns marks.
Not enough time to design, distribute and analyse a survey?
Our marketing research writers handle primary research design, data collection, analysis and full assignment write-up — delivered to your deadline.
How to Analyse Survey Data in a Marketing Research Assignment
Collecting survey responses is the easy part. Turning them into academically credible findings that support your discussion section is where most students struggle. The analysis process below follows the steps expected at undergraduate level — no advanced statistics required.
Enter your data into a spreadsheet
Each row = one respondent. Each column = one survey question. Assign numerical values to all responses — Likert scale answers become 1–5, Yes/No becomes 1/0, MCQ options become numbered categories. Never analyse raw text responses directly.
Tool: Google Sheets or ExcelCalculate descriptive statistics for each Likert question
For each Likert scale question, calculate: mean score (average response), mode (most common response), and frequency distribution (how many respondents chose each option as a percentage). These three figures give you everything you need for your findings section.
Formula: =AVERAGE(), =MODE(), =COUNTIF()Calculate frequency counts for MCQ and demographic questions
For multiple choice questions, calculate what percentage of respondents chose each option. Present the most significant percentages in your findings section — not all of them. Report only what is relevant to your research objectives.
Formula: =COUNTIF() / total respondents × 100Cross-tabulate by demographic variable (if relevant)
If your research question involves comparing responses across demographic groups (e.g. male vs female, under 25 vs over 25), split your data by that variable and calculate separate means or frequencies. This is what produces the kind of insight seen in the Blog 34 example — "female respondents rated peer influence higher at 62% vs 41% for male respondents."
Tool: Pivot table in Excel or Google SheetsCode open-ended responses thematically
Read all open-ended responses and group them into 3–5 recurring themes. Count how many responses fall into each theme. Report the themes and their frequency — e.g. "the most frequently cited barrier was price (mentioned by 28 of 45 respondents), followed by limited product availability (19 respondents)."
Method: manual thematic codingSelect your top 4–6 findings to report
Do not report every question in your findings section — only those directly relevant to your research objectives. Prioritise findings that either confirm or challenge the theories in your literature review. These become the findings you interpret in your discussion section.
Rule: if it does not connect to a theory, cut itSurvey Design Mistakes That Cost Students Marks
Building the survey before defining the research objectives. Students who write questions before establishing what they need to measure end up with data they cannot connect back to their research question. Every question in your survey should trace back to a specific research objective — if it does not, cut it.Fix: Write your research objectives first. Then ask: which question in my survey answers each objective? If an objective has no matching question, add one. If a question has no matching objective, remove it.
Using a scale without a neutral midpoint. A 4-point Likert scale (Strongly Disagree / Disagree / Agree / Strongly Agree) forces respondents to choose a side — which introduces bias by eliminating the genuine middle-ground response. Always use a 5-point or 7-point scale with a neutral option.Fix: Always include "Neither Agree nor Disagree" as the midpoint of your Likert scale. Remove 4-point scales from your survey entirely.
Not anonymising the survey. If respondents know their answers are personally identifiable, they will give socially desirable answers rather than honest ones — especially on sensitive topics like spending habits or environmental behaviour. This is a research ethics issue as well as a data quality issue.Fix: State clearly at the top of your survey: "This survey is anonymous. No identifying information will be collected." Mention this in your methodology section as an ethical safeguard.
Distributing the survey to a convenience sample without acknowledging the limitation. Sending your survey to friends, classmates, or university social media groups is a convenience sample — and that is fine at undergraduate level. What is not fine is presenting findings as though they represent a broader population without acknowledging the sampling bias.Fix: In your methodology, name your sampling method as "convenience sampling" and state explicitly that findings are limited to the sample and are not generalisable to the wider population.
Reporting every question in the findings section. Findings sections that walk through every survey question one by one are unfocused and lose marks for relevance. Markers want to see you select and present only the findings that matter to your research question.Fix: Report only the 4–6 findings most directly relevant to your research objectives. Put complete data tables in an appendix if required.
Frequently Asked Questions
Can I use Google Forms for my marketing research assignment survey?
How do I justify my sample size in my methodology section?
How many Likert scale questions should I include?
Do I need ethical approval for my survey at undergraduate level?
My deadline is in two days and I haven't started primary research — what should I do?
Need Your Marketing Research Assignment Done Right?
From survey design to full analysis and write-up — our expert writers handle every section of your marketing research assignment to first-class standard, delivered on time.
Get Expert Help Today →
