Memory and Cuing Effects on Decision Framing

Amna Kirmani, Duke University
Peter Wright, Stanford University
ABSTRACT - This paper examines how advertising messages can influence the decision criteria people recall in decision framing. Decision framing occurs at the start of a decision episode. In framing, consumers "decide how to decide". One important framing activity is identifying decision criteria, typically from memory. We discuss a person's decision-framing schema and the process by which decision criteria are recalled from it. We propose that ads can influence recall of decision criteria through cuing the activation of the schema during ad processing. In addition, certain message strategies-encourage selective retrieval of decision criteria, while others encourage associative retrieval.
[ to cite ]:
Amna Kirmani and Peter Wright (1989) ,"Memory and Cuing Effects on Decision Framing", in NA - Advances in Consumer Research Volume 16, eds. Thomas K. Srull, Provo, UT : Association for Consumer Research, Pages: 173-175.

Advances in Consumer Research Volume 16, 1989      Pages 173-175

MEMORY AND CUING EFFECTS ON DECISION FRAMING

Amna Kirmani, Duke University

Peter Wright, Stanford University

ABSTRACT -

This paper examines how advertising messages can influence the decision criteria people recall in decision framing. Decision framing occurs at the start of a decision episode. In framing, consumers "decide how to decide". One important framing activity is identifying decision criteria, typically from memory. We discuss a person's decision-framing schema and the process by which decision criteria are recalled from it. We propose that ads can influence recall of decision criteria through cuing the activation of the schema during ad processing. In addition, certain message strategies-encourage selective retrieval of decision criteria, while others encourage associative retrieval.

INTRODUCTION

How do marketing messages influence the way consumers frame buying decisions? Framing refers to cognitive activities, conscious or automated, that structure a decision problem. In framing, consumers deal with five elements of the impending problem- 1) the products and 2) the product attributes (if any) lo be considered; 3) the way attribute levels will be defined; 4) the decision-making procedure or heuristic to use: and 5) situational parameters to take into account. Framing occurs at the start of any decision process, and the initial framing can strongly influence the final buying decision. -

"Frame change" processes have been largely ignored in persuasion research, which has focused instead on attitude change, i.e., how persuasion campaigns influence a person's product-specific beliefs and feelings. Frame change research deals with the process by which communications, decision-making experiences or decision-making environments influence any of the five framing activities cited earlier. Such research deepens our view of behavioral decision-making processes and of the ways in which marketing campaigns affect buying decisions.

In this paper, we study how advertising influences attribute saliences, i.e., which attributes are recalled as decision criteria during initial framing. We focus on overt persuasive attempts to influence attribute saliences, rather than on less direct interventions (e.g., coincidental overlaps among rival ad campaigns, Wright and Rip 1980, or external events that prime particular buying goals, Bettman and Sujan 1987). We also focus on decisions in unfamiliar, complex product categories, in which the subset of attributes people consider may be especially subject to external influence.

We will first discuss the concept of a decision-framing schema and why cuing a framing schema is important in the frame change process. We then discuss factors that affect the relative salience of product attributes as decision criteria, and how various advertising messages may affect this.

THE IMPORTANCE OF CUING A DECISION-FRAMING SCHEMA

A decision-framing schema is a packet of procedural knowledge pertinent to framing activities. This includes knowledge about specific mental operations and larger decision heuristics that could be executed; knowledge about prior framings in the same or related product domains; and expectation-type knowledge about elements of the problem, e.g., about what types of products may be available or about relative attribute importances. A framing schema may exist for a particular product category (e.g., how to decide among fitness centers), at a more general level (e.g., how to decide among complex, risky products), or both.

This type of schema is distinct from a product category or brand knowledge schema, which contains content knowledge about a particular category's (brand's) perceived characteristics, one's utilities for those characteristics, one's global attitude toward the Category (brand), etc.

We propose that changes in framing tendencies only occur when someone's decision-framing schema is activated in memory. In general, then, advertising messages will influence how someone frames a buying decision only if the person activates her framing schema while processing the message. Further, when people process ads they rarely activate their framing schema unless explicitly cued to do so. They are more likely to activate either a brand knowledge schema or an advertising schema, which help them evaluate the advertised brand or interpret ad executions and tactics. These are generally the message processing goals of most consumers. We doubt that people normally activate a framing schema when they encounter an ad because the task of deciding is not immediately salient or because they do not expect to learn much about decision framing from ads.

This theorizing leads to the hypothesis that ads will have a greater influence on attribute saliences in framing when message content explicitly cues a framing schema than when there is no such cuing. Cuing the framing schema means using language or images which draw attention to the concept "me, deciding among products like this" at the outset of ad exposure.

EXPECTED ATTRIBUTE IMPORTANCE AND MESSAGE STRATEGY

One factor that should affect selective retrieval of an attribute and hence its salience as a decision criteria is its expected decision-making importance. Expected decision-making importance is a person's a priori prediction about the probable importance of an attribute in an upcoming decision among brands in a product category. An attribute has high expected importance if a person expects that differences on that attribute among the products to be considered will be strong determinants of her final brand preferences.

We believe that expected importance is encoded in memory as a property of attributes in someone's decision-framing schema (an attribute of an attribute rather than an attribute of a product). We propose that this property guides selective recall in framing; attributes encoded as high in expected importance are recalled first, ceteris paribus. If someone expects to find minor differences among brands, or differences which evoke similar evaluative reactions, she attaches low expected importance to that attribute. If she expects to find attribute levels which she will evaluate quite differently, the attribute's expected importance is high.

When prior decision-making experience has not created clear expectations about attribute importances, marketing communications may exert strong influence on those expectations. Indirectly, ad campaigns may influence beliefs about the levels of an attribute offered by brands in the category or about the utilities consumers should place on those levels. These, in turn, can become inputs into consumers' own judgments of expected attribute importances.

We are concerned, however, with a more direct persuasion process. An advertiser can offer explicit coaching about attribute importance to increase the chance that a target attribute will be treated as an important decision criteria. This coaching should explicitly assert that a target attribute (TA) is important to consider because rival products differ substantially on that attribute in ways that will strongly influence a consumer's personal welfare We call this a Basic Attribute Importance message (Basic, for short).

The coaching in a Basic message will affect the target attribute's salience by influencing selective memory search during framing. However, memory retrieval in framing can also be affected by other factors. Associative network models of memory draw attention to the bonds connecting elements as an important influence on retrieval processes. We reasoned that message impact may be strongest when message content affects both selective recall and associative retrieval processes. So we studied the effects of a Basic coaching message in conjunction with attribute-to- attribute bondings.

We hypothesize that bonding a target attribute to an already salient attribute (ASA), in addition to coaching that the TA's expected importance is high, increases the TA's salience in framing more than coaching without this bonding. The bonding strategy is to assert that the ASA and the TA are both important and should both be considered in decision-making. This type of bonding message may overcome any interference and counter-arguing effects which may undermine a Basic message.

Another type of bonding is between the target attribute and less salient attributes, which may act as decoys. The message might mention one or two attributes, which are not viewed as all that important, along with the TA. The inclusion of these other attributes may result in a part-list cuing effect, so that the salience of the ASA goes down and hence the relative salience of the TA goes up. This effect probably would not hold up over time, since it depends on the recent rehearsal of the part list.

ATTRIBUTE SALIENCE VS. ATTRIBUTE IMPORTANCE IN FRAMING

When someone self-frames a decision, she relies on recall, and our theorizing deals with factors that affect recall processes. A logical measurement, therefore, is free recall of attributes as decision criteria.

This is a different phenomenon from what is measured either by direct questions about attribute importance or by indices of actual attribute importance derived from conjoint analysis. In administering those measures, a researcher frames the decision problem for the respondent, rather than letting the respondent self-frame it. In asking someone to rate attribute importance on scales, the researcher names attributes for the respondent so consider. Likewise, in conducting conjoint analysis, a researcher presents respondents with product portfolios which explicitly name attributes to be considered. In beth cases, the researcher begins by drawing attention to a group of attributes. Any between-attribute differences in salience that existed in people's minds due to ad exposure might be muted by this.

One possibility is that it uniformly eliminates any ad-induced salience differences because people all accept the researcher's framing and do not bother tapping their own framing schema. Another is that all audience members still do their own framing, asking themselves, "Of these attributes they've made me think about, which do I not want to use as my decision criteria?" If so, the prior attribute saliences will determine their responses. A third middle-ground possibility is that some people do one of these and some do the other, which in effect car mute the between-attribute differences due to ad exposure.

Our hypothesis was, therefore, that ad-induced differences in attribute salience which show up in self-framing measures may disappear or be weaker in a measure of actual attribute importance which names attributes at the outset. To test this, we took a conjoint analysis measure of actual attribute importance in addition to the more pertinent free recall measure of attribute importance.

SUMMARY

In summary, the theorizing suggests that ad messages which cue the decision-framing schema, present a basic message, or present an attribute-to-attribute bonding message may increase the salience of an attribute in the decision framing schema. We ran experiments in which Cuing and Message Strategy were manipulated, and attribute salience was measured (via free recall and conjoint analysis). The results were generaLly consistent with the hypotheses.

REFERENCES

Bettman, James R. and Mita Sujan (1987), "Effects of Framing on Evaluation of Comparable and Noncomparable Alternatives by Expert and Novice Consumers,"Journal of Consumer Research, 14 (September), 141 - 154.

Wright, Peter and Peter D. Rip (1980), "Product Class Advertising Effects on First-Time Buyers' Decision Strategies," Journal of Consumer Research, 7 (September), 176 - 188.

----------------------------------------