person Guy Potter access_time August 14, 2019
Survey Jargon For People New To Market Research

Welcome to the wonderful world of marketing, where everything has a name, acronym, jargon or slang. It doesn’t take long to understand, but when you’re adding market research into the mix you’ve got a whole new language to learn!

As a marketer, research is the bedrock to everything you do. Whether it’s to evaluate the success of a campaign, assess product viability or interest, or gain customer feedback, they will become part of the fabric of your role.

We’ve put together a jargon buster to help explain what all these words mean and how they can help you shape current and future surveys.

Survey terms

Margin of error

The expected level of inaccuracy of a survey. In all types of survey, you can be sure that your responses will vary somewhat from the “real” views of the general population.

Only surveying the whole population would give you a zero margin of error. Since that’s not possible, you have to expect and tolerate a certain margin of error.

How much margin of error is tolerable will depend on how accurate you need the results to be – and this in turn depends on the questions and the seriousness of the subject matter. 1000 respondents will yield a +/- margin of error of 3%.


The pool of people who agree to do a survey.


This is a common term that refers to people actually taking part and responding to your survey.

Sample base

Every survey needs a set number of respondents to participate. This number is known as your sample base. Ideally, a survey needs at least 250 respondents to become statistically valid. If you need your margin of error (see above) to be less you may need a larger sample size. This will vary widely depending on the subject matter and demographics of the participants.

If you’re looking for a specific target demographic as well, then the sample base also needs to reflect this. The lower the sample base, the greater the margin of error is likely to be.

Random sampling

An unspecified, randomly selected slice of the population. If you just want general feedback and/or opinion from the public, this is a good way to get a good mix of respondents.

Nationally Representative (Nat Rep)

A survey that reflects the population of the UK, i.e. 51% females / 49% males.

Quantitative research

If you want respondents to answer your survey questions using a yes/no approach/grid questions or choosing from pre-defined multiple-choice or single-choice options.

This gives you the ability to calculate quickly the numbers answering in a certain way. In other words, you can easily quantify opinions using quantitative research. Great for generating graphs!

Qualitative research

Rather opposite to quantitative research, qualitative research helps you to understand your respondents’ thoughts and feelings on a deeper level.

It usually takes the form of focus groups or discussion-based feedback.

Qualitative research and quantitative research often go hand in hand and projects using both approaches are often undertaken by clients , giving you the best of both worlds.

Open-ended question

A question that requires a response or comment from the survey participant who is allowed to use their own words.

When framing open-ended questions, it’s important not to ask a leading question – to allow the respondent to answer as they feel – not how you want them to answer. An open-ended question should not be a leading question.

Leading question

A leading or loaded question is one which implies the answer within or subtly leads the respondent to answer in a certain way. See open-ended question.

Example of an open-ended question: “Tell me your views on a no-deal Brexit”.

Example of a leading question: “What are your views on what we should do to minimise the damage a no-deal Brexit would have on the British economy.”

Closed question

Quite simply, a question that requires a set response. This could be a yes/no response/grid question or single / multiple choice. No narrative answer is allowed.

Research house

The name of a survey or research company.

Types of survey


Usually carried out by a research house, this is a weekly or monthly survey that goes out to a set sample base who respond to a range of topics. As you pay per question, it is one of the more cost-effective ways to do a quick survey. Results are usually delivered within 24 – 48 hours.

Agile research

As the term suggests, agile research is one of the newest trends in market research. It is very flexible and can be carried out quickly and simply. Sometimes in minutes.

Agile research is done online typically using a self-service survey platform. You can either create your own questionnaire or ask the survey house to create it.

You can also select your sample base to be as large or small as you want, with both national or international reach, should you need this.

Once your survey has been created, you simply hit the send button. Responses and reporting will then follow quickly – usually within the hour.

Agile Omnibus

A combination of the above, but think taxi not bus!


Computer Assisted Telephone Interview. Once a key way to do qualitative research, interviewers would complete surveys according to the respondent’s answers. With GDPR now in effect and cold-calling often seen as a nuisance, this type of survey is increasingly being replaced by online and agile research.

Alternatives include CAPI (Computer Assisted Personal Interviews) – where an interview sits (face to face) with a respondent and notes their answers directly on a computer.


When you want to drill down and get a bit more detail from your sample, in order to understand how respondents feel or react to a topic or issue.

Online survey

As it implies, respondents are notified by email to take part and then respond to an online questionnaire. Although web-based, it may take longer than agile research, but these are usually done within 48 hours.

Guy Potter is maru / usurv's head of research and clients.
Recent Articles