Chapter 23: Conducting vox pops

In the previous chapter we looked at this important way of making news into a two-way exchange of information and opinions, between leaders and ordinary people. We considered different types of vox pop, and looked at the advantages of each, for newspapers and for radio and television. In this chapter we look in detail at how to conduct vox pops and how to analyse and present the results.



Whether your vox pop is to be an expensive public opinion survey or a quick street poll, you must prepare if it to be of any use.

What information do you want?

You must have a clear idea of what you are trying to find out. That sounds obvious, but it all too easy to design a questionnaire badly, so that it fails to give you the information which you want.

For instance, if you are conducting a pre-election survey and you want to know who is likely to win the election, you ask people for which party they will vote. If you also want to know the probable shift in political influence, you need to know how these same people voted in the last election. Without this information you cannot do the job properly.

This is where the difference between the types of survey becomes clear. In a structured questionnaire, avoid open-ended questions like "what do you think about..." or "what would you do if..." The answers to these questions are impossible to collate and analyse. You should ask questions requiring yes/no answers, or multiple choice.

For example, a pre-election survey might ask: "Who would you like to be the next Prime Minister?", with the six possible answers: "Julius Chan", "Rabbie Namaliu", "Michael Somare", "Paias Wingti", "somebody else" (asking them to specify who) and "don't know".

For street polls, exactly the opposite applies. What you want is good quotes, and the questions must be open-ended to encourage them. You might ask questions such as "what do you think about photographs of election candidates being included on ballot papers?" or "what do you think of election candidates offering gifts to voters?"

What newspapers can do is to have five or six questions with yes/no answers or multiple choice answers, followed by one or two open-ended questions. In this way, they are able to run a story which has statistical information, but also has some good lively quotes.

For example, your list of questions might be:

  • Will you vote in the next election?
  • Which political party will you vote for?
  • Did you vote in the last election?
  • Which political party did you vote for?
  • Who would you like to be the next Prime Minister? Julius Chan, Rabbie Namaliu, Michael Somare, Paias Wingti, somebody else, don't know.
  • Why should he (she) be Prime Minister?
  • What would you most like the next government to do?

The first five questions will give us results which we can analyse. The last two questions should give us some good quotes.

How many interviews?

You cannot interview everybody: it would take too much time and involve too many reporters, who have other things to do. However, too few interviews will produce results which are unlikely to be representative.

A compromise is difficult but necessary. One good way is to have a team of reporters, each placed at a strategic location in the district on the same day - the market, the shopping centre, the post office and so on. Each one should be able to talk to about 30 people in three hours, as long as the questionnaire is sensibly short. For a street poll you will not need to interview so many.

A target figure for each interviewer gives balance. It avoids the situation where one person interviews only 12 people, another does 25 and a third does 89. This would give undue weight to the views of people in one particular location.

Proof of identification

Have some form of official identification with you, in case you are asked to produce it.

^^back to the top

Conducting vox pops

Remember that the people you wish to interview are under no obligation to answer any of your questions. You are a representative of your newspaper, so be polite at all times.

For each person you interview, you must do these things:

  • tell them who you are, where you are from and what you are trying to do
  • ask for their help and apologise for any inconvenience
  • start a new page of your notebook for each interview
  • make sure that you get their full personal details - name, age, home province, address, job etc
  • if you are asked, try to give them some indication of when the interview might be published; never say that it will definitely be used, as that will only disappoint them later if it is not used
  • take a head-and-shoulders photograph of them

^^back to the top

Analysing the results

With a street poll, it is enough to find the liveliest and best quotes, and use them. However, you will also need to calculate roughly how many were "for" or "against" any particular issue, and choose the quotes so that they reflect that balance as well as giving you the best quotes.

With a questionnaire survey, the analysis is more important and more difficult. You will be faced with a mass of statistical information: what should you do with it?

Count the sample

First, count the number of people interviewed in the vox pop by adding together the number interviewed by each reporter. You can then divide them into different categories, if you wish.

For example, if every reporter makes a note by each interview as to whether it was a man or a woman; and which age group the person was in; then you can divide the sample into useful categories.

Perhaps you had five reporters working on the vox pop, and they each did 30 interviews. In this case, you have a total sample of 150 people.

Add up the number of men and the number of women - it might be 81 women and 69 men. Of course, the numbers must add up to 150.

Perhaps you asked the reporters to put each person into an age group - 18 and under, 19-50, or 51 and over. (You would need to choose appropriate age groups for your society.)

Add up the number of people in each age group - it might be 41 aged 18 and under, 83 aged 19-50, and 26 aged 51 and over. Once again, of course, these numbers must add up to 150.

You could categorise your sample in any way you choose, depending upon the make-up of your society.

Add up the answers

Next, take each question in turn, and add up the answers. For example, for the first question, "Will you vote in the next election?" your answers may be:






This is misleading, though. We know that 41 of our sample were aged 18 or under; they are too young to vote. We need to look at the remaining 109 and add up only their answers. Our answers now are:






You can give this as the answer "from those old enough to vote". You will need to continue, analysing each answer in the same way.

But are we going to ignore all the answers from the people aged 18 and under? We are not. When it comes to the last three questions, their answers will be relevant. Let us take the fifth question as an example: "Who would you like to be the next Prime Minister?" You might add up all the answers and find the following:

Chan: 24 Namaliu: 39 Somare: 22
Wingti: 38 Others: 15 D k: 12

The news angle in this case would be that opinion is evenly divide between whether Mr Namaliu or Mr Wingti should be the next Prime Minister. (It must be stressed that these figures have been made up for this exercise: they are not the results of a real vox pop.)

However, when you break these figures down, according to categories, other interesting things emerge. First, you break them own according to sex:

Chan: 24
11 men
13 women

Namaliu: 39
12 men
27 women
Somare: 22
12 men
10 women
Wingti: 38
19 men
19 women
Others: 15
12 men
3 women
D k: 12
3 men
9 women

You can see from these figures that Mr Namaliu appears to be more popular among women, while Mr Wingti has equal support from men and women.

Now you can break down the figures according to age group (in the following figures Y means young, or 18 and under; M means middle, or 19-50; and O means older, or 51 and over):

Y 0
M 24
O 0
Y 1
M 34
O 4
Y 1
M 7
O 14
Y 28
M 10
O 0
Y 9
M 6
O 0

D k:
Y 2
M 2
O 8

You can see a very interesting pattern emerging from these figures. While support for Mr Namaliu and Mr Wingti is evenly divided, we can see that it is coming from very different age groups. Mr Namaliu is very popular with the 19-50 age group, while Mr Wingti is by far the most popular with the young age group. This is significant for the political future of each man and his party, and therefore for the country, so this would make an interesting news angle for your story.

Historical comparisons

The last thing to do, especially in a vox pop about voting intentions, is to compare the way things are now with the way they used to be. This might mean comparing the results of this vox pop with the results of a previous vox pop; or it might mean comparing answers to different questions in the same vox pop - as with comparing the second and fourth questions in this example.

Perhaps the answer to the second question, "Which political party will you vote for?", was as follows:

National........ 3
Don't know.....7

PAP.............. 4
United.......... 2
Others.......... 9

(Note that the numbers total 92, which is the number of people who said they would be voting.)

We can now compare this with the answer to the fourth question, which asked the same people how they voted at the last election. The result might be:

Don't know...7


Most of our sample seem to be fairly constant in their political support, if these figures are anything to go by. However, PPP's support has dropped by 33%, PDM has dropped 12%, PAP has almost doubled, and support for other parties (and independents) has increased a lot.

This, too, is significant material, and could provide some indication of which parties can expect to do better than last time, which parties can expect to do less well, and which can expect to perform at about the same level.

^^back to the top

Writing the story

Once you have the results analysed, you must treat them like the raw materials for any news story. Pretend that you have just encountered them for the first time; look at them; assess their news value; decide on the most newsworthy angle and make that your intro.

Clearly the story has to be presented to the readers or listeners as a normal news story, written in words. You cannot throw a pile of statistics at them and expect them to sort it all out for themselves.

Nevertheless, once you have presented the most interesting information in a well-written news story, you can give detailed information tabulated at the end. In this way, the readers who are particularly interested can do some of their own interpretation and analysis from your statistics.

As we mentioned earlier, though, do not try to conclude too much from a limited street survey, and even less from a street poll. From an accurate opinion poll you can extrapolate. This means that, for example, if 30% of the women of a particular age and social group interviewed held a particular view, you can confidently say that 30% of all women of similar age and social group hold this view. You cannot extrapolate in this way from our more instant and amateur surveys.

You can say "more than two-thirds of the people interviewed thought...", but you cannot go on to say "most people in Papua New Guinea think..."

^^back to the top

Vox pops for radio and television

Radio and television stations usually only conduct the simplest kind of vox pop - the street poll - with a single reporter asking one or two questions to a small number of people. Its only purpose is to gather a few comments on a current issue, to hear what some ordinary people are thinking. No attempt is made to draw any statistical conclusions from the answers or to suggest that the comments are a guide to how people in society as a whole react. They are a way of illustrating a story, not a way of discovering any new information about what the population thinks about an issue.

Very few radio or television stations attempt the kind of statistical survey you have just seen demonstrated for newspapers. Wealthy television stations may pay a specialist public opinion company to conduct a survey for them, usually at election time. Radio stations seldom pay for such surveys for the simple reason that they cannot use the information properly, as newspapers or television can. Listeners will be bored by a mass of statistical figures.

Conducting a vox pop

Although many of the principles of the street poll vox pop apply equally to both print and broadcast journalists, there are some obvious differences. The most important is that, although television can use film or video of people interviewed, vox pops are useless for radio and television unless you can hear the actual words people use - what we call the actuality. (See Chapter 48: Radio and television basics.) So you must always use a tape recorder or television camera. Before you leave the newsroom, check that all your equipment is working properly.

Choosing a location

Choose a busy place to conduct the interviews, but keep away from loud background noise. For television vox pops the background might be relevant; vox pops about buying habits can be conducted in shopping centres, vox pops about travel at bus stops or train stations.

People are always interested when they see a reporter approach them with a microphone or camera, but some may be shy and try to avoid you. If you let all the shy people escape, you will not get a wide sample of people; you will only get those people who are happy to appear in your report. So try to talk people into giving you "a quick interview". Tell them it will not take more than a couple of minutes. Be polite and if they refuse again, let them go.

You will usually attract quite a crowd, which you can occasionally turn to advantage by bringing in observers for quotes. On television, try to avoid young children standing behind the interviewee, making faces into the camera. Choose a location which will not allow them to do it, for example against a wall or on a piece of high ground.

Using your recording equipment

Once you have them in front of the camera or microphone, help them to relax. If your vox pop is on a humorous subject, maybe relax them with a little joke. Although you must hold the microphone close to the interviewee, do not frighten them with it. Many people are uncomfortable with strange technology like microphones or cameras. So do not push the microphone or camera into their mouth. On television, the camera operator should be a little distance behind the reporter, shooting over the reporter's shoulder.

Asking questions

It is best to ask only one question of everyone. Although it is possible to ask two or more questions, this will create problems later when you come to edit your tape. Decide the exact wording of the question before starting the vox pop. Keep the question short and simple.

Never ask questions to which people can give "yes" or "no" answers. Ask open-ended questions, especially ones beginning with "How", "Why" or "What". For example, instead of asking "Do you agree with the proposed law to ban smoking in public places?" you should ask "What do you think about the law to ban smoking in public places?"

Avoid interrupting their answers unless it seems that they might go on talking for too long, then politely thank them for their time and move to the next interview.

It is not usual for radio vox pops to use people's names, but television reporters might occasionally take people's names to use in captions.

Ending the interview

Your interviewee might want to hear what their voice sounds like, and ask you to replay the tape. Refuse politely, as it will only slow you down. They may also ask when the program will go to air. Because you are not yet in a position to guarantee that their particular interview will be used, do not promise them anything. But tell them approximately when the report will be used. They can then tell all their friends and relatives to tune in at that time, increasing your audience.

How many interviews?

There is no fixed limit on the number of interviews you should conduct. Some of the interviews will be better than others and you may not use some of them at all. If you have only a two-minute space for your report, you may be able to use up to 10 short answers or only four or five longer answers - perhaps an average of 6 voices. It is always better to have some in reserve, so always do a few more interviews than you think you will need. As a rough guide, two or three interviews are not enough, twenty will be too many.

Editing your tape

Whether you report for radio or television, the principle is the same. The quotes must be strong, easy to understand and in good taste, with no swear words or other bad language.

You must listen through the whole of your tape again when you return to the editing suite, making a written note of the best quotes and their location on the tape. If you have had to use multiple questions or a large number of interviewees, make a grid with the interviewees down one axis, the questions down the other. Once you have filled in remarks about each answer, you can look at the grid to choose the best and fairest sample for editing.

You do not need your own questions between the voices of your interviewees. You should write your cue or intro script to explain the issue and what question you asked, but then allow the people's voices to be heard replying one after the other. Edit their comments very tightly, taking out all unnecessary words and pauses.

It will make your tape more interesting if you mix a variety of voices, male and female, young and old, well-educated and uneducated and so on. In multilingual societies, you can also mix languages, although you should check with your producer that this will be all right for the particular program chosen.

Finally, make sure that all the sound is adjusted to the same level. There is nothing more annoying than trying to listen to a bad sound recording.

^^back to the top


Choose which kind of vox pop you are going to conduct according to the needs of your organisation and of the individual story.

If you are conducting a survey for newspapers, make sure that you:

  • ask fair and honest questions
  • ask a representative sample of people
  • record your answers accurately
  • report what they say fairly.

Before you start your survey, you must decide:

  • what information you want
  • how many interviews you need.

When conducting your survey, you must:

  • introduce yourself
  • be polite
  • start a new page in your notebook for each interview
  • get personal details such as name, age, job
  • take a head-and-shoulders picture of them.

When analysing results, you should:

  • Count the sample
  • add up the answers
  • compare the figures against each other.
  • Treat the results of your analysis like any news story.

When conducting vox pops for radio or television:

  • prepare a simple, open-ended question
  • make sure your equipment is working well
  • choose a good location
  • be friendly and relaxed with the interviewees
  • interview enough people for your needs - but not too many

When editing your tape:

  • choose strong comments
  • do not keep repeating your question
  • mix the voices to give variety
  • make sure all your sound levels are correct.

^^back to the top

>>go to next chapter


Index to Chapter 23
  1. Preparation
  2. Conducting vox pops
  3. Analysing the results
  4. Writing the story
  5. Vox pops for radio and television
  6. To summarise
Looking for something?

A quick way to find what you're looking for in The News Manual is through the Index. It has more than 900 links to concepts throughout the manuals. Click here:IndexLink0

QuickLink to introductions of The News Manual volumes

Home | About | The Manuals | Exercises | Resources | Links | Contact Us | What's New

Copyright David Ingram and the Peter Henshall Estate 2008. Website by Diopdesign