Skip to main content
Search form
  • 00:00


  • 00:11

    DESIREE LOPEZ: My name's Desiree Lopez,and I'm the CEO of TNS BMRB.We're one of the UK'S largest social research agencies,and we've been around for over 80 years.We started life as BMRB, which stands for the British MarketResearch Bureau.And we became part of TNS in 2008.We have a very wide portfolio of studies that we undertake,

  • 00:32

    DESIREE LOPEZ [continued]: both for government departments and academic departments,and, also, third sector organizations.We do everything from collecting data for national statistics.So we collect all of the data for the British Crime Survey,right through to doing policy studies for Whitehalldepartments.That can include formative studiesto be able to understand what kind of policy

  • 00:53

    DESIREE LOPEZ [continued]: should be generated, right through to impact studiesto try and evaluate whether or not policy's been effective,both from a policy articulation point of view,but also trying to understand the impact of new policieson the general public.We also have a specialist team that evaluates governmentbranding communications.And that includes looking at government behavior change

  • 01:14

    DESIREE LOPEZ [continued]: programs.Social research is many things, but in general, it's reallythe understanding of people.What we need to be able to do is understand the way peoplebehave, their attitudes, their perceptions about the world.We need to understand the interrelationshipbetween, for example, policy, and how people experience

  • 01:35

    DESIREE LOPEZ [continued]: policy in the real world.And that could involve, for example,the way in which changing the rules around taxation impactspeople's behaviors, or changing the regulations around accessto benefits actually impacts people's lives.So I head up a team that conduct surveysfor government and public sector clients

  • 01:57

    DESIREE LOPEZ [continued]: within TNS, the wider company.So we help our clients either make decisionsby providing evidence or evaluate whether the policiesare working.So when they spend a lot of moneyon a program for people seeking work, for example,they need to know if it's working or not.And we do that research to help them find outwhether it's being effective.So to describe what a survey interview is

  • 02:20

    DESIREE LOPEZ [continued]: could be many different things according to howwe conduct the interview.So as a survey company we not onlydo interviews face to face with people in their home,but a lot of interviews are conducted over the telephone.Or we might send a postal questionnaire to the respondentto fill in, or, increasingly, these days, peopleconduct surveys online.So how it feels for the respondent

  • 02:41

    DESIREE LOPEZ [continued]: is going to vary a lot, accordingto how we administer the survey, most conventionally, a faceto face interview.Our interviewers will go into the house with a laptop,and as researchers will have designed a questionnaire that,that interviewer will administer and make surethat the respondent is completely assured

  • 03:02

    DESIREE LOPEZ [continued]: about the confidentiality of the survey.None of their individual results will go to the clientand that they really understand the purpose of the research.I'm thinking particularly about our government work.It really is important to instillthe importance of the research in respondents,and how it's going to be used to inform decision-making.Hello, my name's Chris Phillips from TNS.

  • 03:22

    DESIREE LOPEZ [continued]: That's the largest research company in the country.And I've been given very few addresses to go to.One of them is this.And I'd really appreciate it if youcould help me with my survey.[INTERPOSING VOICES]It's very important.And this is a little thank you leaflet that we giveto everybody that we speak to.And it will reassure you that we comply with data protection.

  • 03:45

    DESIREE LOPEZ [continued]: And it will tell you something about the company we work forand the standards and rules of confidentiality.The times that it's right to do a surveyare when you really want to know how many people are doinga certain thing, behavioral measurements, what people'sattitudes are to something.For example, we measure how many people

  • 04:05

    DESIREE LOPEZ [continued]: are concerned about climate change.The government needs to know how much interest there is in that.And, also, we measure a lot of satisfactionwith public services, so customer satisfaction levels.So surveys give you those kind of statisticsto understand how many people are doing something,or how many people are thinking something.They often work hand in hand with qualitative research

  • 04:25

    DESIREE LOPEZ [continued]: because once you found out how many people are doingsomething, your next question is often, well,why are they doing that?And that's where we combine the datafrom quantitative and qualitative researchto really get under the skin of the findings.So there's two ways of conducting an interview.You can conduct a quantitative or a qualitative interview.

  • 04:46

    DESIREE LOPEZ [continued]: A quantitative interview will have a very pre-setquestionnaire that the researchers have designedin conjunction with the client, where you're most often askeda question and you have to select from a list of setanswers what your response is.We'll sometimes ask what's called an open-ended question,and the interviewer will record a sentence or two.

  • 05:08

    DESIREE LOPEZ [continued]: So that's a quantitative-- that's a survey interview.A qualitative interview will go into even much more open-endedremit.They'll have a topic guide of the subject matterthey want to cover, but the respondent can take themin many different directions.It's a conversation.And, then, the researcher-- well, the interviewer--who is the researcher in this case,

  • 05:28

    DESIREE LOPEZ [continued]: will go back and analyze the text of that conversationto find patterns and meaning from that interview.There's a whole range of research methodsthat are available in the tool kit for social researchers.If we think about qualitative methods, everything from focusgroups to depth interviews, observations, ethnography,

  • 05:50

    DESIREE LOPEZ [continued]: deliberative techniques.And, of course, with technology really, being able to help usunderstand people's experiences.Ethnography, the ability to do online communitywork, the ability to actually usefilm in terms of how we're observing and also collectingdata.And which of these food issues, if any,

  • 06:10

    DESIREE LOPEZ [continued]: are you concerned about?Please select all that apply.OK, I'll put the food poisoning and hormones and [INAUDIBLE]in food, food hygeine eating out,pesticides, and food hygiene at home.

  • 06:31

    DESIREE LOPEZ [continued]: It's on the side.And [INAUDIBLE]?No, I think that's about it.OK.The design stage of a survey could be very involvedwhen you're working with government and public sectorclients, because it's really importantto get the measures right.So our clients will come to us with an idea of whatthey want to understand.But it's our job as researchers to design a survey that's

  • 06:51

    DESIREE LOPEZ [continued]: going to measure the things that they want to know about.We often find that when we come upwith a first draft of the questionnaire,that helps our clients refine their researchaims in the first place.And it becomes a very iterative and collaborative processof refining and tweaking that questionnaireto really get to the heart of whatthey're trying to understandBut even when we got to that stage

  • 07:12

    DESIREE LOPEZ [continued]: and we think we have a perfect questionnaire,we move on to what's called cognitive testing stage veryoften.And this uses the discipline of psychologyto conduct a very special type of interviewwith a small number people, maybe, a dozen or so people.Where, for every question that we'regoing to administer in the surveywe ask the question, and then we ask them

  • 07:34

    DESIREE LOPEZ [continued]: what question meant to them.And we conduct more qualitative style interviewto explore what processes were going on in their mindand what they were thinking of when we ask the question.That normally requires, then, another redesign stage,because it challenges our thinkingabout whether that questionnaire was ready to go outinto the survey.

  • 07:54

    DESIREE LOPEZ [continued]: And we go back to the client and refine the questionnaire again.We then move into another stage of piloting the questionnaire.And, again, we need to check it's working.Because when a government client, particularly,is spending millions on a survey,it's absolutely vital to get it right.Social research is an incredibly exciting placeto be because it's always changing.

  • 08:15

    DESIREE LOPEZ [continued]: So, although, we're interested in some very core themes,health, education, welfare, crime, the core issues,the things that are most important,the things the government wants to pay attention to,the things that are most important from the communityperspective, those change all the time.

  • 08:37

    DESIREE LOPEZ [continued]: And so, it makes a job, and a jobthat you can have, really, over a very long career, alwaysexciting and always diverse.What's also, of course, very exciting isthe way in which methodology is changing, the way in whichtechnology, the way in which being more mobile,the way in which things like social and open datais being integrated into much more

  • 08:59

    DESIREE LOPEZ [continued]: traditional social research studies, more traditionalsocial research methodologies.So it's enabling us, really, to potentially evengo back to a question that may now have changedover a very long period of time, but, actually,having to rethink our designs, rethink the way in which weanswer that question.Because we now have a whole host of data available to us

  • 09:20

    DESIREE LOPEZ [continued]: that we may not had in the past.So types of surveys we do for the government and other publicsector bodies cross all the different policy areasthat you can imagine.So one of our flagship projects is called the "Crime Surveyfor England and Wales," and that'sused by the home office and many other government departments.It's well known because it comes up

  • 09:40

    DESIREE LOPEZ [continued]: with estimates of crimes that do not depend on people reportingit to the police.So we'll ask people about their experience of crimeand use those to produce statistics that can be comparedto police recorded crime.But we also do a lot of work for the Departmentfor Work and Pensions, who obviouslyinvest a lot of money in employment schemes,

  • 09:60

    DESIREE LOPEZ [continued]: back to work schemes.And that's the evaluation part of our work, wherewe will conduct surveys amongst peoplewho have taken part in some programsthat they've been running.And we'll track them over time, and go back to themand see if there's been any impact of that policyon their job prospects.We also are very well known in TNS BMRB for conducting what's

  • 10:22

    DESIREE LOPEZ [continued]: called longitudinal research.And these could be some of the most fascinating studies.So these studies are designed to track and understandchanges in society over time.The survey we do for the Department of Educationis called a longitudinal study of young peoplein England, where we select people-- young people,when they're 14.And we go back to them every year

  • 10:43

    DESIREE LOPEZ [continued]: or so, through to their early 20s.And it's an absolute goldmine of information for departmentto understand what the circumstances, whatfactors in your adolescence impact on your eventual career.In home interviews, when we conduct faceto face interviews, having arranged an appointment that'sconvenient for the respondent, because often

  • 11:05

    DESIREE LOPEZ [continued]: these interviews are going to take 30 minutes, 40 minutes,or so.And it needs to be fit around their daily life.We'll try and find a quiet space.Something that's very important to considerwhen conducting a survey interviewis in-home confidentiality, as well.Because, for example, if you're talking to a young person, doeshe really want their parents to hear the reportsabout their smoking and drinking behavior, for example.

  • 11:26

    DESIREE LOPEZ [continued]: So we'll either to do that, but, also,when we're designing a survey with these laptops in the home,we'll also designed certain sections of the surveyto be able to turn around to the respondentand ask them to fill it in.And that's what we use for recordingsensitive information.So exactly how it goes is at the discretion of the respondent,where they want to make space.And the interviewers are very adept at standing

  • 11:47

    DESIREE LOPEZ [continued]: in hallways, if they have to in orderto get that interview done.All right, which of these describe your place?I am married.Thank you.And we'll just take that one back.And which of the following apply to you?

  • 12:07

    DESIREE LOPEZ [continued]: None of them.None of those.No.And which of these best describe your ethnic group?White British.The nature of the interview will vary very much accordingto the mode of data collection.If you're in the home, then, you can settleinto quite a long interview.We can design some of our most complex surveys that way.And it can take anything up to an hour of questioning.

  • 12:31

    DESIREE LOPEZ [continued]: And the interviewer will, hopefully,be invited into the home and have the laptop.And the questionnaire is programmedinto special survey software, ready for the interviewerto administer.In a quantitative survey, it's very importantthat, that question is read out exactly as it is on the screen.And our interviews are trained so that the question must be

  • 12:51

    DESIREE LOPEZ [continued]: repeated exactly as designed.Because. otherwise, you would collect inconsistent dataon a very grand scale.The same principles apply to telephone interviewing.We need to make sure that the question is read out exactly asintended.Online, we monitor a little bit about how participantstake part, how long they take to conduct the survey,

  • 13:13

    DESIREE LOPEZ [continued]: whether they sort of dropped out halfway through,got bored or anything like that.And, similarly, postal surveys, it'svery important to design an extremely simple form.We'll think we can understand questionnaires.But, actually, keep it absolutely simplebecause we can't follow instructions very well.One of the key factors that we're always

  • 13:34

    DESIREE LOPEZ [continued]: considering when designing a surveyis how long that interview's going to be.And there's no hard and fast rule about this.It doesn't have to always be under 20 minutes, or so on.Some of our surveys are pretty long, can be over an hour long.But respondents need to find it really interestingand engaging.So you need to look at the subject you're researching,and really think about what's called respondent burden.

  • 13:57

    DESIREE LOPEZ [continued]: Can they sustain a conversation about this subjectfor as many minutes as you need to be in the home?And that's where the questionnaire design stage willgo through many iterations of really honing down to,what are the essential questions we needed to ask?Because, otherwise, most policy questionnaireswould be about two hours long, because there's lots of things

  • 14:17

    DESIREE LOPEZ [continued]: we'd like to ask.But it would be too much for the respondent,and our response rates would go downbecause people would drop out of the interview.So we're now going to ask you a few questionsfrom another subject.OK.Which.If any of these gambling activities have youspent state money on in the last [INAUDIBLE]?Well, I would say none, unless the lottery's there.

  • 14:42

    DESIREE LOPEZ [continued]: [INAUDIBLE]No.So when we've completed a survey, and thatcould be 1,000 interviews, or 20,000 interviews,depending on what we're looking at.So the data all comes down the wire, it's back to the officeand into statistical software that we use to analyze data.At a very basic level, we give our clients

  • 15:05

    DESIREE LOPEZ [continued]: descriptive results.We just describe what percentage of the populationare doing a certain thing or thinking a certain thing.So, for example, in a survey we do for Sport England,we measure the levels of exercisethat different types of people take.But of more value to our clients isthe exploratory and explanatory analysis

  • 15:25

    DESIREE LOPEZ [continued]: that we do to try and understand,well, why are certain types of people doing something,and why are others not?So that's where we build statistical models.And we put in lots of factors that mightbe affecting that behavior.We hypothesize about what is it that drives a certain behavior.And we find statistically significant associations

  • 15:47

    DESIREE LOPEZ [continued]: that we can explain to the client.These are the things that might predict someonewho's likely to take exercise.And that could be really useful for policy-making to think,well, these are the things we needto target to help people achieve the goals that wewant to achieve.When analyzing the data, It's important to havea very clear plan of what you're going to do.

  • 16:08

    DESIREE LOPEZ [continued]: We conduct huge surveys, and you couldbecome overwhelmed by the data.And that's where, again, it comes backto being a good communicator with the clientand going right back to the original aims of the research.What were they trying to do?It's very easy to drift off into all sortsof interesting avenues that mightbe interesting to explore.And I've worked in academic sector

  • 16:28

    DESIREE LOPEZ [continued]: where you might have more time to do that.But in a policy environment, we need to get to the answer.We need to have something that could be taken to a ministerto brief them on what is driving this behaviorand what are our recommendations in terms of what to do next.So I think keeping a very clear head and sticking to the plan.[MUSIC PLAYING]

Video Info

Publisher: SAGE Publications Ltd.

Publication Year: 2017

Video Type:In Practice

Methods: Data collection, Survey research, Qualitative interviewing, Longitudinal research, Questionnaire design

Keywords: adolescence; branding; change detection; crime; customer satisfaction with services; effectiveness; employment; exercise (physical activity); food; government; government agencies; inconsistency; internet; investment; public policy; technology ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



TNS BMRB is a social research agency that conducts studies for government agencies, academic departments, and third sector organizations. The studies they do involve survey interviews. Because interview style can affect the end results, researchers have to take into account where to do the interview, how to maintain confidentiality, and how to design the questionnaire.

Looks like you do not have access to this content.

Data Collection and Survey Interviews: TNS BMRB

TNS BMRB is a social research agency that conducts studies for government agencies, academic departments, and third sector organizations. The studies they do involve survey interviews. Because interview style can affect the end results, researchers have to take into account where to do the interview, how to maintain confidentiality, and how to design the questionnaire.

Copy and paste the following HTML into your website