SMS Polling: A Methodological Review

1
SMS Polling A Methodological Review

• Joel Down
• MORI

2
Overview

• Why SMS polling?
• Research design and methodological issues
• Summary and applications

3
Why research via SMS?

• Increasing penetration of text messaging
• Opportunity for interaction with TV audience
• SMS reaches parts other methodologies cant

4
Mobile surveys reach places others cant
Q Where are you completing this survey?
Office
Lounge
Bed
Travelling
Base All WAP users (387), Sept 2000
5
Survey objectives

• Deliver data for use in BBC Watchdog broadcasts
• Assess representativeness of SMS polling
• Identify strengths and weaknesses of methodology

6
Issue 1 sampling

• No single source of mobile phone sample
• Difficult to collect demographic data
• Opt-in requirements
• Cost for participant
• Recruit panel members through BBC Watchdog

7
Recruiting SMS users
Watchdog call to action
Text address to BBC
Complete web form
Complete postal qre
Collate profile data and phone numbers
Watchdog SMS panel
8
Population and panel profiles
100
0
GB Adult Population
Watchdog Viewers
SMS Panel
9
Population and panel profiles
100
0
GB Adult Population
Watchdog Viewers
SMS Panel
10
Issue 2 question design

• SMS length
• Response format
• Text versus precode
• Scales versus single code

11
Issue 3 speed of response
Base All respondents (1,811), January 2003
12
Issue 3 speed of response
Base All respondents (1,811), January 2003
13
Issue 4 data processing and analysis

• Cleaning data
• Matching data
• Weighting data

14
Issue 5 validation of findings
Q How satisfied or dissatisfied are you with
the service from your main bank?
SMS Survey
MORI Benchmark
Satisfied
Neither/nor
Dissatisfied
Dont know
Base All respondents (1,567/12,000)
15
Issue 5 validation of findings
Q Do you think it is safe or unsafe for parents
to choose the MMR vaccine for their
children?
SMS Survey
MORI Telephone Omnibus
Safe
Unsafe
Dont know/not stated
Base All respondents (1,814/1,008)
16
Summary

• Constraints
• Sample - access and representativeness
• Question length and format
• Cost of participation
• Data processing and analysis

17
Summary

• Opportunities
• Increasing penetration of non-voice services
• New messaging technologies
• New billing structures
• Access to hard-to-reach groups
• Speed of response

18
Thank youjoel.down_at_mori.com020 7347 3000