Summary of methodology
OBJECTIVES:
The objective of the survey was to gather information on the radio listening habits of the South Africans over a 7-day period.
More specifically, the objectives covered:
• Radio stations listened to during a 7-day period
• Times listened to each station, for each day of the week, for each quarter hour of the day from 00h00 to 24h00
• Places where radio was listened to "yesterday"
• Radio stations listened to in the past 4 weeks
• Favourite radio stations
UNIVERSE:
New Populations:
• The latest 2011 Census data from StatsSA was used by IHS to update and release new mid-year 2013 population estimates. These new population figures have been incorporated from the November 2013 SAARF RAMS release.
• The resultant population updates are different to the previous estimates for a number of demographics, changing population profiles.
• The new population update has set a new benchmark and therefore figures are not comparable prior to November 2013.
• Minority sub-populations previously excluded from the AMPS/RAMS universe are now included. This is the sixth RAMS release with 100% coverage (previously 98.5%). The only significant change is in the Black sample outside Cape Town. As a result, the Western Cape coverage is not comparable with surveys before November 2013.
• The +/- percentages and thousands for the Aug 2014 audience figures may only be used in comparison with the June 2014, May 2014, February 2014, December 2013 and November 2013 releases.
The survey covers adults aged 15 years and older, who are living in the nine provinces of South Africa.
METHOD:
Radio self-completion diaries were placed with respondents to complete in their homes over a 7-day period, for which they were offered a small incentive.
Respondents were instructed on how to complete the SAARF RAMS Diary.
SAMPLE DESIGN:
A probability sample stratified by area was drawn.
The sample was equally apportioned between males and females with the exception of mines, hostels and domestics. At each sampling point a cluster of two addresses was drawn.
In each household, the main respondent was selected by means of a Random Grid. Four calls were made before substitution could occur.
During the survey period, household "flooding" was implemented. In addition to placing a diary with the main respondent, supplementary diaries were placed and completed by all other members of the household aged 15 years and over.
As of 2009, the flooding methodology was expanded to include the small urban and rural areas.
The main advantage of flooding is to allow for reporting on a more robust sample. A total of 25,624 diaries were collected over the two fieldwork waves.
FIELDWORK:
The large urban fieldwork for this third release wave of 2014 was conducted in 2 parts: the first part from mid-March 2014 to the first week in May 2014, and the second part from the first week in May 2014 to early July 2014.
The small urban/rural fieldwork for this release was conducted from mid-January to end June 2013. This is the same small urban/rural database that has been used in all the RAMS releases since November 2013.
The current SAARF RAMS contract was awarded solely to Nielsen.
Fieldwork since 2009 has been conducted solely by Nielsen.
A 20% back-check was conducted on all interviewers.
ANALYSIS OF RESULTS:
All findings in the Radio Services report are based on adults 15 years and over.
Reporting is on the combination of the two waves of interviewing. Results were weighted up to total population.
Analysis of results was provided in total, and by demographic breakdowns as follows:
• Age
• Community size
• Sex
• Home language
• Household income
• Household purchaser
• Languages understood
• Languages spoken (Other)
• Level of Education
• Living Standards Measure
• Metropolitan area*
• Population group
• Province
• Mothers' status
• Work status
*Since the November 2013 release, East Rand and West Rand codes are available on the data file.
In addition to the trended figures, the latest RAMS release also includes two additional columns:
a) After the % column (reflecting % of adults who listen to a particular station) is a +/- % column which indicates the possible range of this audience %, with a statistical certainty of 95%.
b) After the '000 column (reflecting a station's audience in thousands) is a +/- '000 column which indicates the possible range of this audience figure, with a statistical certainty of 95%.
Special Note regarding Western Cape Community Stations:
The Voice of the Cape concluded its shared broadcast with Radio 786 on 26 August 2013. These two radio stations are now broadcasting on separate frequencies. As a consequence, for the May, June and August 2014 releases, there are 3 separate lines of data available in this Radio Services Report:
• National trended data for The Voice of the Cape combined with Radio 786
• The Voice of the Cape large urban data only for February, May, June and August 2014
• Radio 786 large urban data only for February, May, June and August 2014.
These same three datasets will also be available on the August RAMS release data files.
