Evolution in research

The EMRO conference, attended by 57 media audience researchers from 20 European countries, was hosted and organised by the South African Advertising Research Foundation (SAARF), the only non-European media research body invited to be a member of this organisation when SAARF was formed in 1974.

Two of the key media audience research issues discussed at the conference were global trends in media audience research and data fusion.

Piet Smit says: "The media audience research industry has been struggling with data fusion for some time now. Originally, researchers worked with classic fusion, where you would merge two databases to create a third database from which to draw your information and then do your cross-tabbing and other analyses from the latter source."

Telmar UK's Dick Dodson gave an overview of newer methods, such as 'data fusion on the fly' and multi-basing, which are gaining increasing attention.

"These methods are once-off fusions, which you do for a specific target market or product," explains Smit. "For instance, you would link variables from two datasets which were pertinent only to beer drinkers. You would get different results when you fused for wine drinkers. These are good methods, but do pose the risk of producing anomalies where the credibility of your data could be called into question."

The issue of mixed methodologies also received a lot of attention at the conference. Part of the need for mixed methodologies is the cost of face-to-face interviewing, and the increasing difficulty of gaining access to certain target markets.

"In South Africa, our biggest problem is the upmarket person, who is very busy, and very security minded," says Paul Haupt. The obvious solution would be to reach these groups via other channels, such as an Internet or email questionnaire.

Using different methodologies within one survey however, is not as simple as it may sound. Each methodology will produce different audience levels: "In a face-to-face interview for instance, the respondent may over-claim his reading of glossy magazine titles to impress the interviewer. In an emailed questionnaire, where there is no interviewer, you could get very different consumption levels."

The Dutch are already using the Internet for the bulk of their interviewing, and use face-to-face for the small percentage of respondents who don't have Internet access. This topic was again placed on the agenda for next year.