SINGLE-SOURCE - NEW ANALYSES

More information is available from single source panels than is often realised. This includes descriptions of the category and of the brand, and the associations between brand share and price, demographics, wei ght of viewing and recent viewing. Data can be aggregated into weeks a nd normal time series modelling compared with the disaggregate finding s; the latter seem to be the more sensitive. Reasons for the brand cho ice at each purchase occasion can be studied by multivariate regressio n. These include the shopper's loyalty to the brand, its relative pric e, a trend term, and recent advertising for the brand and for its comp etitors measured by adstock. Short-term advertising effects have been seen at two to 28 days half life for various brands; no effects have b een found for some others. Competitors' adverse effects may be larger or smaller than ours. Diminishing returns to higher current advertisin g pressure can also be measured and are usually slight. A minority of occasions are under high pressure, and most of these are for heavy vie wers who are also affected by competitors' activities and often have u ntypical brand shares. Any bivariate relation between recent advertisi ng exposure and brand choice is potentially affected by purchase/viewi ng bias, which often occurs, and by other confounding factors, such as price. Such relationships can give misleading indications of advertis ing effects.