Survey’s broadcast results do not mirror the technical realities of coverage

Fri Mar 6, 2015

Editor: I read with interest your analysis of the latest media poll by Systematic Marketing & Research Services and the obvious conclusion that the construct and execution of the survey are faulty.{{more}}

I have had the opportunity to analyse the electronic media results up to 2011 and found anomalies that do not support the accuracy and integrity of the results.

For example, the survey stated in 2011 that its active audience was 78,100 people over the age of 14, but 2007 estimates from the Statistical Department showed that the total number of Vincentians in that category should have been 69,507. The active audience was hence 12% higher than what should have actually existed. It was also 14% higher than the available audience that should have followed the Census 2001 figures. This anomaly only affected net audience volumes, but not placements.

The survey sample in 2011 was also biased towards the 35- 54 age groups with over sampling in the 35- 44 category by 5%, and by 9% in the 45 – 54 category as well, when compared to 2007 population estimates. In using the 2001 Census figures, the bias was worse with under sampling in the 15 – 19 age group by 11% and over sampling by similar margins in the 35 – 44 and 45 – 54 categories. This anomaly severely affected stations that specifically targeted the 15 – 25 age group.

Looking back at Systematic’s 2010 Survey Field Plan, it was observed that there was a significant bias away from the Windward population sampling towards Kingstown and the Leeward coast to a lesser extent. The Census population divisions reported that 51% of the population resided on the Windward side of the island from the Calliaqua division northwards, inclusive of Marriaqua. Kingstown residents made up 12% of the population while Leeward and the Grenadines made up 29% and 8% respectively.

The survey’s field plan showed that only 32% of those polled resided on the Windward side, while 28% resided in Kingstown. The Grenadines sample was consistent with documented population geographic spread but the Leeward sampling was over reached by 3%. Hence the surveys have been focusing their poll on Kingstown and the Leeward side of the island. So effectively, Systematic polled 16% more Kingstown residents and 19% less Windward residents.

The results then just did not mirror the technical realities of coverage. For example, the NTRC can confirm even today that only NBC can be heard island-wide. Hitz FM reaches Owia and Nice Radio reaches as far as Sandy Bay. Hot FM, Star FM, and WEFM cover the Leeward side but fade out and in, past Georgetown. While Hot FM may have and could still be number one, it was and still is inconceivable that it could have polled so high overall.

To top it all off, in 2015, a station that has been off the air for more than three years, Total FM, still showed up in the results with a 1% audience and brought to mind the old adage, “Statistics are comfort for a fool.”

Broadcast Analyst