Summary Knowing More Than We're Told: Inferences From Advertising Claims

Carolyn J. Simmons, Lehigh University
Gita Venkataramani Johar, Columbia University
[ to cite ]:
Carolyn J. Simmons and Gita Venkataramani Johar (1994) ,"Summary Knowing More Than We're Told: Inferences From Advertising Claims", in NA - Advances in Consumer Research Volume 21, eds. Chris T. Allen and Deborah Roedder John, Provo, UT : Association for Consumer Research, Pages: 96.

Advances in Consumer Research Volume 21, 1994      Page 96



Carolyn J. Simmons, Lehigh University

Gita Venkataramani Johar, Columbia University

Advertising presents products in the most favorable light, emphasizing positive attributes and describing them in ways that maximize their attractiveness. In turn, consumers bring their expectations, knowledge, and information needs to bear on their interpretation of ads. Inferences are an important part of this interpretive process. This session examined the inferences that consumers make in response to common advertising practices and the impact of these inferences on subsequent judgment and behavior. Taken together, these papers (a) contribute to work on contingent models of inference making, recognizing the moderating roles of involvement and knowledge, (b) recognize a range of consequences of incomplete or ambiguous information, including effects on attitude confidence, attribute importance, and product category perceptions, and (c) address methodological issues relevant to understanding when inferences are generated.

The first paper by Sandra Burke, Sandra Milberg, and Youjae Yi examined the effects of claims about attributes which are inherent to the product category, but which are not emphasized by competitive brands (e.g., Mazola corn oil has no cholesterol). Among those who believed the claimed attribute to be atypical of the product category, such claims increased the probability of brand choice and the perceived importance and typicality of the attribute. These subjects also reported believing that brands not making the claim did not possess the attribute.

The second paper by Gita Venkataramani Johar primarily examined claims with incomplete comparisons (e.g.,this brand is better) and inconspicuous qualifications (e.g., qualifications in small print). The inconspicuous qualification was deceptive only under low involvement. Deceptive inferences from incomplete comparisons appeared to be made by some low and high involvement subjects; however, response times revealed that such inferences were made at the time of processing the ad only under high involvement.

The third paper by Frank Kardes examined the effects of set size, valence of information, and instruction set (encouraging consideration of unmentioned attributes or not) on consideration of unmentioned attributes, inferences, brand evaluations, and confidence and uncertainty about evaluations. Positive information increased positive inferences, while negative information increased negative inferences, with the greatest number of negative inferences being made when subjects were encouraged to consider unmentioned attributes. No effects were found for set size. Positive inferences had a stronger impact on evaluations than did negative inferences, whereas negative inferences had a stronger impact on confidence than did positive inferences.

In his concluding discussion, John Lynch addressed methodological issues relevant to the measurement of deception, including designs for determining whether an inference not made at encoding is made at the time of choice, the tradeoffs in using response times versus closed- and open-ended questions to detect inferences, and the use of baseline control conditions to detect inferences. He also identified a number of questions raised by the papers regarding potential mediators (attribute diagnosticity, inferences versus discounting, etc.) and moderators (sparseness versus denseness of information environment) of observed effects.