Getting to the truth in readership data

Getting to the truth in readership data THE SUN THE reawakened media debate on the issue of reliable newspaper readership figures has called for clarifications and cor rections as media owners race each other to argue over methodology and accuracy of each other's findings. In a statement ,the Joint Industry Committee for Print Audience Research (Jicpar) said "There is a need for maturity and rationale in all for us to find a common ground to end this dispute". Nielsen Media Research (NMR)'s face-to-face interviews for its Media Index feature about 10,000 individuals aged 15 years and above in a year and the results are reported quarterly based on 12-month rolling sample. The media coverage as reported by NMR has been the sole cur rency used by the advertising industry but media owners have raised their discontent as they find their expanding readership or coverage had not been reflected in Nielsen's results. It is felt that the main reason for this was that interviewers have not sufficiently reached individuals living in guarded homes, apartments and condominiums ?these are the individuals with diversified media habits. As such some publications with upscale readership profiles felt to be under-represented here and this has led dailies to commission Synovate and TNS, which have produced findings disparate to those published by NMR. These were expected as different research methodologies and survey period (one off against rolling 12 months) were used. The Jicpar board and the technical committee reviewed the Nielsen Media Index research methodolog y in February. While Nielsen claims to adopt most of the international guidelines for readership research, it agreed to an independent audit to verify its methodology and implementation. Jicpar has recommended that Nielsen over-sample the urban areas to create an effective robust sample of key advertisers' target for analysis and profiling. Before reporting the results, such over-sampling should be downweighted to reflect the correct proportion in the total population. Recognising the under coverage of individuals living in guarded homes, apartment and condominiums, Nielsen has used numbers extracted from the telephone directory to recruit the sample from condominiums and luxury apartments. When Jicpar last checked, the achievement rate for these dwelling, although improved, was still low. In July, Jicpar invited Nielsen, Synovate and TNS to propose the best research methodology that can overcome the above limitations. Despite the limitations, Nielsen maintains that faceto-face interviews are the best data collection method for readership research in Malaysia and proposes to reweigh the sample to overcome the limitation. While face-to-face interviewing is the most widely used method for readership survey, there is a shift to use CATI (Computer Assisted Telephone Interviews) interviews internationally to reach affluent individuals. Synovate and TNS have shown that it is possible to reach individuals living in gated communities and condominiums using the telephone. They proposed mixing both methods of data collection to overcome the limitations. For the rural area, they proposed only face-to-face interviews. Synovate and TNS agreed that face-to-face and CATI can be combined effectively to boost response rate and reduce costs, as long as researchers minimise the dif ferences between the questionnaires and samples and monitor the profile of responses. The mixed mode sampling and weighting procedures should be checked for validity. Synovate provided Jicpar the example of its findings compared to the NMR numbers and it was seen that phone interviews resulted in higher numbers for higher income adults with higher readership across the board. Jicpar, which is composed of the Malaysian Advertisers Association (MAA), the Association of Accredited Advertising Agents Malaysia (4As), the Media Specialists Association (MSA) and newspaper media owners, is not fully convinced that a combination of face-to-face and phone interviews will result in more accurate numbers. At this stage, Jicpar is not endorsing one methodology over the other. Synovate has launched its Media Atlas in Malaysia. Media Atlas is very different from the Media Index, as the former proposes to significantly over-sample urban. Nielsen practises proportionate sampling to cater to many different categories of users. The information collected is also quite different. The additional readership survey will help the industry to verify the methodology and hopefully resolve the controversy in readership numbers. In the long term, the industry can move to one or keep both. It is Jicpar's intention, however, to work with the research companies to ensure the final accuracy of their numbers. Jicpar's main goal is to secure readership data that are reliable and a methodology which is relevant, transparent and effective. Jicpar has appointed a consultant and is likely to work in the area of audit with NMR. In due course, Jicpar intends to make a definite recommendation to the industry Jicpar chairperson and president of the Media Specialists Association, Margaret Lim, says "It is our belief that the existence of two currencies will act as a check and balance and they will be competitive. It is up to the media planners to trust their judgment and consider the figures that have been made available, which would only raise the planners' effectiveness. We have world class research companies but face an evolving lifestyle, where the higher income consumers are becoming more security conscious, too busy and elusive. Methodology which worked 30 years ago may not be as effective today. Jicpar acknowledges that while face-to-face interviews have difficulty in penetrating the af fluent homes, there are also limitations in phone interviews. In the long term, when the digital era has grown in full and we have achieved high inter net penetration, webbased interviews for the future must be seriously considered. Media companies now are more research-savvy and, we hope, the industry will be able to work with Jicpar to ensure the accessibility of accurate and reliable research data. Jicpar will continue to work with all parties to find the best solutions for the industry. We have to be realistic that no one party will be able to come out with a figure fully acceptable and we implore every industry player for its participation and co-operation on this matter". Issued by Jicpar Kuala Lumpur