“Sixty-three per cent of respondents would recommend the canteen.” Next to it: “Feedback about the canteen was mostly positive.” Both sound like findings, yet only the first comes from a quantitative survey and only the first can be checked.
A quantitative survey is a standardised piece of research: every respondent receives the same questions, in the same order, with the same answer options. That deliberate equal treatment is what produces data you can count, compare and analyse statistically, even across several hundred people. By the end you will know the four common modes of data collection and the point at which a large number of responses says surprisingly little.
📌 Key points at a glance
- Standardised means identical questions, order and answer options.
- Four modes: online, paper, telephone, face to face.
- The ONS Annual Population Survey reaches roughly 80,000 households.
- Taking part in ONS household surveys is voluntary.
- Without random sampling, findings describe only the respondents.
Create a survey for free
With empirio.ai you can create a modern online survey in minutes — with hosting in the EU.
- AI-built survey
- Adjust by drag & drop
- Real-time analysis
What is a quantitative survey?
A quantitative survey is a study in which a questionnaire fixed in advance is put to every respondent in exactly the same form, so that the replies can be expressed as numbers and analysed statistically. Methods literature calls the same thing a standardised survey; both labels mean one and the same approach.
The word “quantitative” refers to the form of the data, not to how many people took part. A study of 40 people with fixed answer options is quantitative. A conversation with 500 people in which everyone speaks freely is not. Treating “quantitative” as a synonym for “lots of people” confuses the outcome with the method.
Standardisation is not a formality. It makes sure that differences between the answers come from differences between the people, rather than from one respondent having been asked a differently worded question. Only on that basis may you calculate means, compare groups and test relationships.
Quantitative describes the form of the data, not the number of respondents.
A quantitative survey does not have to consist purely of closed questions. A few open fields are normal, typically a comment box at the end. Once the open part carries the actual research aim, however, the study is qualitative and a different set of rules applies to the analysis.
The four types of quantitative survey
Quantitative surveys differ by the route the questionnaire takes to reach the respondent. That route is technically called the mode of data collection. Four modes are in common use, and large studies routinely combine them rather than settling on one.
All four rest on the same standardised questionnaire, with a fixed question order and binding answer options. That shared basis is what later allows results collected by different routes to sit in a single dataset.
| Mode of data collection | How it works | Best suited to |
|---|---|---|
| Online survey | questionnaire in a browser, self-completed | large groups, short fieldwork, small budgets |
| Paper survey | printed questionnaire by post or in person | groups without reliable internet access |
| Telephone survey | interview by phone following a fixed script | random samples drawn from phone numbers |
| Face-to-face survey | interview on site, usually on a laptop | long questionnaires, complex topics |
The Annual Population Survey shows how this works at scale. According to the Office for National Statistics, each annual dataset achieves a sample of roughly 80,000 households, or about 175,000 respondents, collected through computer-assisted face-to-face and telephone interviewing (as of August 2026).
Taking part is voluntary, and that is a British particularity
The Office for National Statistics is explicit that nobody has to take part in the Labour Force Survey if they do not want to. Several other European countries handle their equivalent survey differently: in Germany, Austria and Switzerland the corresponding official household study carries a statutory duty to respond. The British approach places the burden on persuasion instead, which is why response rates are watched so closely by the ONS.
Why online is now the default
The online survey has become the norm because it combines three things that used to work against each other: short fieldwork, no interviewer costs and answers that already exist in digital form. Routing and consistency checks run automatically, catching mistakes that on paper only surface during analysis.
The price of going online: the coverage problem
One drawback is methodological and weighs more than it first appears. Bandilla (2015) sets out the coverage problem in the GESIS Survey Guidelines: there is no complete list of people with internet access and no list of email addresses from which a random sample could be drawn. Telephone numbers can be generated at random; email addresses cannot.
Designing the questionnaire: what actually matters
A quantitative questionnaire stands or falls on its answer options. The question itself is usually written quickly; the real work sits in the categories the respondent has to choose between.
An uncomfortable rule applies here: a questionnaire does not only measure, it also influences. Bogner and Landrock (2015) summarise in the GESIS Survey Guidelines how strongly design and ordering shape answering behaviour. Three of their findings can be applied straight away.
The order of answer options shifts the result
When answers are shown visually, on screen or on paper, respondents pick one of the upper options more often. This primacy effect reverses as soon as the options are read aloud: on the telephone the last options named move ahead, which is the recency effect. The same question can therefore produce two different rankings in two modes. Short answer lists and rotating their order keep the effect in check.
Agreement scales invite agreement
Acquiescence is the tendency to agree with statements regardless of their content. Citing Krosnick, Bogner and Landrock (2015) draw an awkward conclusion: this bias can only be prevented reliably by avoiding agreement scales such as “agree / disagree”, “true / false” or “yes / no” altogether. Asking about frequency or asking for a rating gives the tendency far less to work with.
A midpoint is a decision, not a default
Where a rating scale offers a midpoint, that category is chosen disproportionately often, including by people who hold a clear view. The phenomenon is known as the tendency towards the middle and is a form of satisficing, the mental shortcut. None of that makes a midpoint wrong: where genuine indecision is plausible, leaving it out forces respondents into a position they do not actually hold.
💡 Tip
Ask three people from your target group to complete the finished questionnaire while thinking aloud. Ambiguous wording shows up in those 20 minutes far more reliably than in any silent proofread.
Benefits and limits of a quantitative survey
The main benefit of a quantitative survey is comparability. Because everyone is asked the same thing, results can be aggregated, broken down by group and held against earlier waves.
The limits sit in exactly the same place. Anything the questionnaire does not provide for never appears in the data, and nobody misses it, because the analysis can only show what was asked.
What a quantitative survey delivers
- Answers stay comparable across every respondent.
- Large groups are reachable at reasonable cost.
- The analysis is transparent and repeatable.
- Results can be tracked across several points in time.
What you accept in return
- Reasons and nuances stay outside the answer options.
- Mistakes in the questionnaire cannot be fixed after launch.
- A questionnaire needs more preparation than an interview guide.
- Numbers look more solid than they often are.
The last of those limits carries the greatest risk. A percentage always looks precise, no matter how the respondents ended up in the study. Whether a finding holds beyond the people who answered is decided by the sampling method rather than by the number of replies, and what follows from that is set out in the article on representative surveys.
Create a survey for free
With empirio.ai you can create a modern online survey in minutes — with hosting in the EU.
- AI-built survey
- Adjust by drag & drop
- Real-time analysis
Quantitative or qualitative: which fits your question?
The choice between quantitative and qualitative does not turn on effort, it turns on how your research question is phrased. If it asks how many, you need numbers. If it asks why, you need accounts.
A simple test helps: can you write down the possible answers in advance without leaving out anything important? Then a quantitative survey is right. If you would be guessing what people might say, a qualitative pilot belongs first.
| Your question is | Suitable method |
|---|---|
| How many students use the library each week? | quantitative survey |
| Why do some students avoid the library? | qualitative interviews |
| Has satisfaction changed since 2024? | quantitative survey, repeated |
| Which reasons have we not thought of at all? | qualitative pilot, then quantitative |
The bottom row describes how good dissertations usually proceed: a handful of open conversations first, to learn what the possible answers even are, then the standardised questionnaire that makes them countable. The full comparison of both routes is in the article on qualitative and quantitative surveys.
Common mistakes in quantitative surveys
Four mistakes turn up again and again in dissertations and in club or team surveys. All four can be sorted out before launch and none of them afterwards.
What stands out is that none of them has anything to do with statistics. They all happen at the planning stage, long before the first response arrives.
Mistake 1: mistaking a lot of responses for representativeness
A survey shared on Instagram and through your own mailing list reaches the people who happen to hear about it and fancy taking part. Bandilla (2015) states plainly that self-recruited groups of this kind cannot be used to estimate population parameters. In plain terms: 800 responses describe those 800 people, not “students” in general. That is no disaster, but the methods chapter has to say so.
Mistake 2: treating the mode as a question of cost
On sensitive topics people answer differently depending on who is listening. Bogner and Landrock (2015) report that socially desirable answering weakens when social distance is large, as in a self-completed online questionnaire, and strengthens in a face-to-face interview. For a survey about drinking, exam anxiety or conflict in a team, the mode is therefore a methodological decision rather than a logistical one.
Mistake 3: asking too many questions
Every additional question costs concentration, and at some point answering behaviour tips into satisficing: the respondent only skims and picks whatever seems plausible. Bogner and Landrock (2015) name three factors behind it, and two of them are in your hands, namely how difficult the task is and how motivated respondents are. A short questionnaire with clean questions beats a long one with many.
Mistake 4: starting without an analysis plan
Before the first question goes into the questionnaire, it should be clear which analysis that question will later support. An age question in five bands allows no mean; a yes-or-no question allows no gradation. Anyone who thinks through the methods of analysis only after fieldwork regularly finds that the one variable that mattered is missing.
Conclusion
A quantitative survey is the right tool as soon as you want to count and already know which answers are possible. Its strength, standardisation, is also its limit: it delivers clean numbers for exactly the questions asked beforehand and for no others. The effort therefore belongs before launch, not after it.
Where to go next
- Ready to build the questionnaire? How to create a questionnaire
- Still choosing a method? Qualitative and quantitative research methods
- Looking for the bigger picture? Empirical research: definition and guide
Ready to run your own quantitative survey?
With empirio.ai, an online survey tool from Germany, you set up the standardised questionnaire, share it as a link and watch the analysis build while responses come in.