This paper describes a panel management strategy using targets for household composition, including some related to rapidly changing Population characteristics. These targets may be used to control panel imposition through a stratification scheme or may be used to weight Panel results to avoid bias. Typically, targets include characteristics such as subscription to pay TV services or the number of television sets. The method is primarily used in TV rating panel operations involving TV establishment surveys as an integral part of a two-phase design, implying he selection and survey of a master sample determining strata targets at |d a classified sampling frame to select panel homes. This design is Rudely used but the methods for determining targets and recruitment rom survey respondents are frequently inefficient and result in biased v audience level estimates. A simple construct is proposed to overcome such problems, involving classification of survey respondents and panel homes both by characteristics at the time of the survey as well as at the hue of each rating report.
In this paper we describe in detail the Australian panels which source the AGB BrandScan consumer purchasing information service. In describing these panels we give a flavour to the way in which the shape and nature of the BrandScan panels were determined. In later sections we will address some particular issues in methodology which have a bearing on our fundamental thesis that all data capture instruments should in their design be acutely sensitive to the total impact, both perceived and real, on the selected respondent. In a panel context the use of the 'new technology', carefully specified and implemented, should not only increase the level of accuracy in reporting but also give rise to higher co-operation levels than those obtained on panels using traditional diary techniques. From here it is a relatively short step to argue that a panel with a low data capture burden associated with its normal task can be used to provide high quality information on - say - TV viewing, albeit for restricted periods of time, particularly if appropriate technology is used.
This paper describes the UK Electricity Council's use of panel research techniques for load forecasting and energy marketing. In particular it details how panel data are used to help in forecasting the shape of the domestic component of electricity demand and explain changes in the level and shape of household demand.
This paper has been prepared by senior executives at both AGB and Quantime, who, during the first six months of 1981 collaborated very closely throughout the specification, program writing and implementation phases of a computerised despatch/booking-in system for the Home Audit panel. In the first section we review the history of AGB's Home Audit panel. In subsequent sections we look at current operational aspects and outline possible future developments.