How to write surveys - practical guidance

skills for planning, designing, and writing surveys

What are surveys?

Surveys are usued to gather statistical data, first-hand evidence, or customer feedback and opinions.

Surveys are commonly used to gather information about

voting preferences during elections
government public health surveys
market research on commercial products
public opinion on controversial social changes
population surveys and census taking
data for an academic project or test

Surveys are often a form of specialist product research. They are often created by someone who has been trained in market research.

Graduates in Economics, Psychology, Sociology, Mathematics, or Business Studies are often asked to construct and evaluate the results of a survey as part of their coursework.

However, people working in other disciplines are increasingly asked to write surveys as part of their work.

Surveys – the process

If you attempt your own survey, you need to be aware of the procedure outlined here in this example.

Establish the goals of the project – What you want to learn
Determine your sample – Whom you will interview
Choose interviewing methodology – How you will interview
Create your questionnaire – What you will ask
Pre-test the questionnaire, if practical – Test the questions
Conduct interviews and enter data – Ask the questions
Analyze the data – Produce the reports

Surveys – Example

Statement of objectives

I intend to find the main reasons why Mighty Box is more popular than Pink Bucket as a gift container in the north west of England.

Rationale

The results of this survey will inform the marketing team regarding the next advertising campaign due to start 18 May 2014.

Methodology

I will carry out my research by distributing survey sheets to 5000 consumers.

I will create a database which will analyse my results in terms of a stated hierarchy of preferences and reasons.

Implementation

My consumer profile comprises women aged 25-50 in social groups C2 and D.

Location

My survey will be carried out in 17 shopping centres in the north west.

Human resources

I will employ experienced market research personnel in each of the target areas.

Costings

17 researchers @ £7 per hour x 18 hours = £ 2,142.00

Database technician @ £150 per day x 2 days = £ 300.00

Analyst @ £20per hour x 40 hours = £ 600.00

Total cost £ 3,042.00

Surveys – the two main types

There are various purposes in carrying out a survey. Here are the two main types, categorised by the types of question posed – open and closed.

Open survey
Closed survey

Open questions – What are they?

Open questions are those that allow the respondent to make up their own answer and express it in their own words.

Here are some examples of open questions.

Describe your feelings about the rail-crash.
What is your current opinion of the NHS?
How do you deal with distressed patients?

Closed questions – What are they?

Closed questions have a prescribed answer, as in multiple choice questions. Yes or no answers follow closed questions.

Here are the same topics expressed as closed questions.

Do you feel angry, sad or depressed about the rail crash?
Do you agree that the NHS is declining in quality?
Are you able to control distressed patients?

When are open questions more appropriate?

In a survey, open questions yield more authentic opinions and therefore can widen the scope of a survey. This is because the participant may express ideas that you have not thought of.
If you are genuinely attempting to find out what a group of people do think or feel, open questions are effective.

When are closed questions more appropriate?

When you are seeking to categorise a set of known symptoms, behaviour, beliefs, or feelings, closed questions are more efficient.
When you are seeking a certain response, to a given idea, then closed questions can assist your project.

Surveys – Open

Enlightenment

This is an ‘open survey’ – one for which there is no preconceived notion of result.

In other words, you need the result to be as authentic as possible, otherwise it will have no value. You have no vested interest except to know the facts. You will use these facts to inform your next sales strategy or publicity campaign.

Constructing the survey content needs extremely careful thought, planning, and trialling. An inefficient survey can cost you dear.

The open survey. To find authentic information, questions of a very different order are required. Here are some examples of open questions.

Q: I see you have bought a Mighty Box. Please state in one sentence your reason for choosing it.
A: I chose the Mighty Box because it is easy to carry and it is suitable for boys and girls.

Q: Indicate how often you read women’s magazines.
A: Never – every week – every month – every three months.

Q: Tick the content of women’s magazines in order of your preference.
A: None – articles – stories – adverts – horoscopes – readers letters

Q: With which of the following magazines are you most familiar?
A: Hers – Glamour – Lifelong – Woman’s Groan – Modern Girl – none of these.

Surveys – Closed

Publicity and politics

This is a ‘closed survey’ which seeks to gather the information which will support a belief or a statement of belief. In this case you know at the outset what result you are seeking.

For example, if you are a toothpaste company and you want to use statistics in your advertising campaign, you need to find those people who already use your product and are happy with it.

If you want to convince your management team that the main office is in need of refurbishment, you might construct a survey which will give you results to convince them. In this case, you might even include some Health and Safety issues to create a more powerful effect.

The closed survey demands ‘closed questions’. These are most likely to give you the responses you seek. Here are some examples of closed questions with their predicted answers

Q: How long have you been enjoying Mr. Stipling’s cakes?
A: Ever since they came out in 1976.

Analytical statement: ‘X number of people stated that they had enjoyed the cakes for many years’.

Q: When you relax and read Woman’s Groan, what do you like best about it?
A: I only read Woman’s Groan for the horoscopes. I don’t read the rest of it.

Analytical statement: ‘In a recent survey, over X thousand women said they found Woman’s Groan relaxing.’

Surveys – Guidelines for user-friendliness

1. Decide at the outset which of two methods you will use:

hand over the questionnaire for each subject to complete
read out the questions and fill in the results yourself

2. If you hand over the questionnaire, you need to –

Make all your questions simple and clear
Make the text large enough to be legible in the street
Make the questions answerable with a tick or one word

3. If you decide ask the questions, you need to –

phrase questions so that they sound like a spoken sentence
keep the survey out of sight of your subject, so that they actually listen rather than read over your shoulder

4. Always explain the context and the purpose of the questionnaire.

5. Avoid including the individual’s identity, or if you need it in some cases, ask permission to include it.

6. Always thank your subject for co-operating with your research.

Processing your data

7. Creating a database is the best way of dealing with survey results. This way, you can find a variety of results which will be done automatically, saving you hours of manual labour.

8. A database will also help you to generate results that you may not have anticipated at the outset.

Using the data

9. Use the data to inform your activities in as many ways as possible. The authentic survey is extremely valuable for such purposes as

correcting an ineffective publicity strategy
checking on the efficacy or usability of a product
assessing public relations in your company
using ideas generated by your employees
checking on employee morale

10. When acting on the results of your survey, publish your results as an integral part of your action statements.

11. For example, if your survey has shown that employees are not happy with the level of Health and Safety in your organisation, state that you are responding to their input into the survey.

12. If, as a result of the survey you need to make changes to your product or service, give your employees the benefit of the survey results.

Expressing the results of your survey

13. Charts, diagrams or tables are better than narrative explanation when it comes to expressing survey results.

14. A graph or bar chart showing preferences makes a bigger impact than a page of written text on the subject.

15. There are many easy-to-use electronic applications for creating graphics such as those needed to express statistics. They work in conjunction with databases.

16.A spreadsheet such as Microsoft Excel can output its data in the form of bar-charts, columns, and pie charts.

More on How-To
More on literary studies
More on writing skills