tracking research of public information campaigns in the ... · pdf filequestionnaires...
TRANSCRIPT
Tracking research of publicinformation campaigns in the
Netherlands
Marc Arnold, Jaap Veenstra
Presentation setup• The Netherlands Government Information Service• Public information campaigns• Why tracking research?• Production process before use of SAS
• Tracking Data Warehouse• Production process with the use of SAS
• Questions
Department of General Affairs
The Netherlands GovernmentInformation Service/DTC
Division of communication research
Public information campaigns
Organisation• 25 - 30 campaigns per yearCharacteristics• Use of massmedia, advertising and marketing
techniques• Own brand: Postbus 51• Accountably to the parliament (meta-evaluation)
• Individual campaigns:– problems with former research design
• New media, and more old media ($ 30 million)• Shift in attention for cost efficiency• Standardised research focused on:
– effect off campaigns– efficiency off media use
• More users off information:– campaign managers– media managers– policy makers– government cost efficiency
Why tracking research?
• For each campaign– pre measurement (n=600)– weekly survey of media reach (n=1200)– post measurement (n=600)
• Continuously– media use– appreciation Postbus 51 brand– Political issues
Research design tracking
• High standards– telephonic screening: stratified random sampling– face-to-face interviews– use of multi media laptops (CAPI/CASI)
Questionnaire
Media consumption
Political issues
Media reach & appreciation
Background variables
Process dataOther
Awareness
Attitudes
Knowledge
Behaviour (intention)
Remembered reach
Reach specific media
Appreciation campaign
Functioning campaign
Campaign-specific4 or 5
DTC-intake
questionnaires campaign material
Screeningsurvey
Preparation field work
executionField work
data filesDTC
Process tracking research
Primary process
Secondary process
Primary production process:reporting
Voor-meting
Bereiks-meting
Effect-meting
Recognition
0
10
20
30
40
50
60
70
80
90
wk 01 wk 02 wk 03 wk 04 wk 05 wk 06 wk 07 wk 08
%
campaign recognition
TV spot
radio one
radio two
outdoor
ad's in newspaper
SPSS code SPSS output
To what extent are you interested in......n ot tt
in ter es ted n eu tr al in ter es ted% % % befor e after
total 19 47 34 63 .5 4 .3*
gen der **men 17 47 36 3 .8 4 .4 *women 30 52 18 43 .4 3 .7 *
age **18-24 34 53 13 3.2 3.325-34 22 48 30 03 .9 4 .1 *35-49 27 45 28 64 .0 4 .3 *50> 19 40 41 44 .1 4 .4 *
r epon den t is a **smoker 42 34 24 43.9 4.1non-smoker 18 48 34 54 .6 5 .0 *
r each ed **reached 19 43 38 0- 4.5not reached 32 53 15 5- 3.9
aver age on7-poin t s cale
metaevaluation
Analysingtrends
BrandawarenessPostbus 51
Secondary production process:reporting
Political issues week 1 - week 33
3,5
4,0
4,5
5,0
5,5
6,0
6,5
7,0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
safety/criminality
education
environment
Dutch economy
taxes
Building expertise
• What can we expect off a communicationcampaign in terms off effects?
• Is there a relation between media pressure andeffectiveness off campaigns– What’s the most efficient media mix
• in general / specific issues• What’s most important:
– a good concept / use off media• Building simulation models
Old
pro
duct
ion
proc
ess
DTC-intake +questionnaire
questionnaires + show material
screeningsample survey
Preparation field work
executionField work
data filesDTC
Process tracking research
Primary process
Secondary process
DTC-intake +questionnaire
questionnaires + show material
screeningsample survey
Preparation field work
executionField work
data filesDTC
Process tracking research
Reporting
new
pro
duct
ion
proc
ess
Database
Reporting
DTC-intake +questionnaire
questionnaires + show material
screeningsample survey
Preparation field work
executionField work
data filesDTC
Process tracking research
Database
Consultancy
•New Name
•New Skills & Competencies
•New Profile and Image
•New Company
eXplore your potential to grow
Tracking Data Warehouse
• Goals– Improve efficiency standard analyses en reporting– Facilitate secondary analyses
• Requirements– Flexibility in analyses– High demands on layout reports
• Approach– Data model– Extraction, organization and exploitation module– User Interface
Former analysis process
SPSS files
SPSS output
Excel tables
1 SPSS file
n ot totalin ter es ted n eu tr al in ter es ted n
% % % bef or e aftertotal 19 47 34 6813 .5 4 .3 *
gen der **men 17 47 36 3333 .8 4 .4 *women 30 52 18 3483 .4 3 .7 *
age **18-24 34 53 13 373.2 3.325-34 22 48 30 1803 .9 4 .1 *35-49 27 45 28 2164 .0 4 .3 *50> 19 40 41 2474 .1 4 .4 *
r epon den t is a **smoker 42 34 24 4183.9 4.1non-smoker 18 48 34 2594 .6 5 .0 *
r each ed **reached 19 43 38 309- 4.5not reached 32 53 15 358- 3.9
av er age on7-p o in t s cale
-Manual data processing-Visual validation-Unpractical categorization
-Cut and paste of statistical output
Building the Data Warehouse
SPSS files
Excel tables
n ot totalin ter es ted n eu tr al in ter es ted n
% % % befor e aftertotal 19 47 34 6813 .5 4 .3 *
gen der **men 17 47 36 3333 .8 4 .4 *women 30 52 18 3483 .4 3 .7 *
ag e **18-24 34 53 13 373.2 3.325-34 22 48 30 1803 .9 4 .1 *35-49 27 45 28 2164 .0 4 .3 *50> 19 40 41 2474 .1 4 .4 *
r epon den t is a **smoker 42 34 24 4183.9 4.1non-smoker 18 48 34 2594 .6 5 .0 *
r each ed **reached 19 43 38 309- 4.5not reached 32 53 15 358- 3.9
av er age on7-p o in t s cale
Database
EXPLOITATIONEXPLOITATION
EXTRACTIONEXTRACTION
ORGANIZATIONORGANIZATION
Metadata
Metadata
Maintenance
Extraction
background survey answers
-Respondent id-Date of birth-Gender...
Exploitation: example of analysis
• Campaign: smoking prevention• To what extent are you interested in ….
• Calculate frequencies en averages• Analyse by gender, age and education• Extra analysis by smoker/non-smoker• Test for differences in pre and post measurement• Test for differences within groups
To what extent are you interested in ….
To what extent are you interested in ….
Secondary analysis: example• Analysis: political issues for wk1-wk33• Explore pattern for issue Education
3,5
4,0
4,5
5,0
5,5
6,0
6,5
7,0
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
safety/criminality
education
environment
Dutch economy
taxes
Conclusions• Time-saving primary process:
– no manual data processing– no cut and paste of statistical output
• Improved quality– automatic data validation
• Meet requirements for– flexibility– demands on reports
• Categorized data suitable for secondary process
⇒ Improved utilization of Tracking Research
Visit our stand for demo and moreinformation