TEACH IN AGENDA

1
Response Ratesand why YOU should be worrying
about them
Richard Asquith 10th April 2001
2
(No Transcript)
3
The most important issue facing market and media
research
4
What do we mean by Response Rate

• The survey universe, for example
• all adults in GB
• adults living in London and the SE
• 15-34 year olds
• Representative sample of eligible respondents
• Response rate proportion of those we attempt to
  contact who we successfully collect data from

5
What Causes Non-Response

• Not available to make contact with
• Refuse to participate
• Provide unusable data

6
Why is it so important?
7
Why is it so important?

• A low response rate increases the risk of
  interviewing an unrepresentative sample
• An unrepresentative sample will yield
  unrepresentative data
• Unrepresentative data will
• distort audience measures
• lead to wrong business decisions

8
A Fictitious Example

• Cars, Gadgets and Gratuitous Pictures of Scantily
  Clad Women
• (Preferably Ones Who Appear on TV)
• Men Women All Adults
• Actual readership 90 10 50
• Survey sample size is 1,000 adults
• Successfully interview 500 500 1,000
• Actually tried to contact 4,000 1,000 5,000
• ie response rate 12.5 50 20
• Readers 250 50 300 50
  10 30

9
A Hidden Problem

• Not possible to calculate the response rate on
  most surveys
• quota samples
• necessary data not collected
• However, the problem will still be there

10
The EvidenceGovt Survey Response Rate Trends
11
The EvidenceThe National Readership Survey
12
The EvidenceGB TGI
13
The EvidenceJICMARS - The National Medical
Readership Survey
14
Response Rate InfluencesGrove Couper 1998

• Social context
• Respondent characteristics
• Interviewer attributes
• Survey design

15
Response Rate Influences Windle Mouncey 1997

• Willingness to participate based primarily on
  seeing usefulness of resultant data
• Either for themselves or society as a whole

16
What can we do about it?
17
What Can We Do About It?

• Interviewer training
• Fieldwork management - eg time of day
• Good survey introduction

18
What Can We Do About It?

• Ensure that respondents see the value
• Appropriate incentives
• Marketing of the MR industry

19
What Can We Do About It?

• Make it easier to participate by providing data
  collection options
• self-completion
• online
• Re-contacts / access panels

20
What Can We Do About It?

• Make it a better experience for respondents
• Reduce the respondent burden
• Data fusion

21
What does this all mean for Media Research?
22
Declining Response Rates Affect Us All

• If we dont address the problem
• poorer quality data
• bad business decisions
• If we do
• costs will inevitably increase
• users must be prepared to accept solutions such
  as data fusion