hierarchy-of-effects

Definition:
A term summing up the proposition that ADVERTISEMENTS exert their influence on the audience by a simple hierarchical progression of effects. First postulated 70 years ago, it continues to dominate the conceptual frameworks of textbook authors and advertising practitioners alike, despite severe criticism on theoretical and experimental grounds over the years. It is a clearly detectable implicit assumption in much if not most of what is written on both advertising effect and the measurement of ADVERTISING EFFECTIVENESS today. Over the intervening seven decades, a considerable number of 'verbal models' of the hierarchy-of-effects have been published in the academic and practitioner literature- each different, but all clearly closely related. The most familiar four are: (1) Daniel Starch, 1923: 'To be effective, an advertisement must be . . .seen - read - believed - remembered - acted upon' (2) E.K. Strong , 1925: 'AIDA': attention - interest - desire - action (3) Robert C. Lavidge and G.A. Steiner, 1961: 'Hierarchy of Effects': awareness - knowledge - liking - preference - conviction - action. (5) Russell H. COLLEY, 1961: 'DAGMAR': unawareness - awareness - comprehension - conviction - action. Psychologists would recognise these as a specific example of the generic 'cognitive-affective-conative' or 'C-A-C' pattern of response to stimuli other than advertisements. Cognitive responses are the outcome of thinking about what is happening, affective responses result from an emotional reaction to the stimulus, and conative responses involve consequent actions. Colloquially, the C-A-C model is summed up by the vividly explanatory 'think-feel-do'.
Though Lavidge and Steiner coined the phrase 'hierarchy of effects' it is nowadays used as a generic term. Strong's AIDA is the most widely quoted of all, despite the antiquity which is seldom made explicit. Very similar models were proposed by E. M. ROGERS in 1962, to explain the adoption of innovations and William J. McGuire in 1969, to explain 'information processing'. To facilitate discussion, let us combine these vriants into the consolidated model of advertising effect below. The levels of response have been arranged in a vertical hierarchy and assigned labels mostly taken from the established versions. The two sets of labels take different perspectives on the process by describing respectively what the advertisment 'should achieve' and how the audience 'should respond'. For example, it should communicate and the audience should comprehend. These two modes have historically been mixed together in a single model. The third column relates the new scheme to the generic C-A-C model in its colloquial form. What the hierarchy-of-effects hypothesis does not do is explain how the audience is propelled, or voluntarily progresses, through the levels of the hierarchy of effects. More specific criticisms have also been made by academic theorists over the last 30 years. Most influential among the objectors is KRISTIAN S. PALDA, who published a widely reported evaluation of Lavidge and Steiner's model. The first of his fundamental objections, based on a priori reasoning, is that progression from one rung of a hierarchical ladder to the next does not mean that the probability of eventual action has necessarily been increased. His second is that in particular circumstances, such as impulse-buying, the deliberate step-by-step progression implied by the hierarchical models may actually be highly telescoped. The third is that he could find no conclusive evidence in the literature to show that 'affective' change (conviction, sympathy) necessarily preceded behaviour change, rather than resulting from it. This calls into question the very sequence of the hierarchical levels, and is certainly the most important of the three. Furthermore, Professor A.S.C. Ehrenberg and his colleagues at the London Business School (and latterly at South Bank University, London) have, over a period of more than 20 year, published the findings of meticulous empirical research studies and theoretical analysis supporting Palda's proposition that attitude change may follow new behaviour rather than causing it. Their view is that, once the decision to try a new product has been arrived at, probably somewhat arbitrarily, and provided that the first trial is not an unsatisfactory experience, a stable pattern of subsequent re-selection develops. The user then deliberately pays attention to advertising for the product, which in turn reinforces the choice. In other words, 'do' responses trigger 'think' responses and precipitate 'feel' responses aimed at rationalisation of a choice already made. This is certainly not the sequence of the orthodox hierarchies. In 1973, Michael L. Ray suggested that three arrangements of the hierarchy were possible. 'Learning' corresponds to the conventional hierarchies: think-feel-do. 'Dissonance attribution' postulates an exactly reversed sequence, which he believes will apply when the decision is an 'involving' one and differences among the brands available are small. The eventual choice (do) is made on a relatively trivial factor, attitudes are rearranged (feel) in order to reduce post-decision dissonance and only finally is close attention paid to the advertising (think) in search of a rationalisation for the choice. A 'low involvement' hierarchy, think-do-feel, applies when the audience is casual about the whole business. Perceptual defences are lowered, permitting awareness and recall to be achieved by sheer weight of advertising, brand choice is then made on the basis of what is best remembered, and attitudes are thereafter steadily rearranged in order to confirm the choice. In 1965, the highly influential theorist Herbert Krugman had argued that a low-involvement hierarchy best described the circumstances of television advertising. Despite the serious shortcomings of the conventional hierarchy-of-effects models, we have to recognise and accept that they are still the implicit conceptual underpinning of present-day advertising practice, in the great majority of cases. In particular, this means that the measurement of ADVERTISING EFFECTIVENESS is implicitly based on a learning hierarchy. This must remain the case, of course, until marketing academics are able to produce a better model which practitioners can understand and are willing to use.

Cross-References:
[AIDA] [Colley, Russell H] [DAGMAR] [Rogers, EM] [Palda, Kristian] [advertising effectiveness] [advertising effectiveness]

Links:

Figures: