The Impact of Framing, Anchorpoints, and Frames of Reference on Direct Mail Charitable Contributions

ABSTRACT - Direct mail has long been an important tool to solicit charitable contributions. While consumers have become more comfortable using direct mail, they are deluged with ever more frequent direct mail solicitations. Direct marketers also have become increasingly sophisticated, which makes it difficult for traditional charitable direct mail appeals to cut through the clutter. The purpose of this paper is to examine three framing-related creative strategies in a charitable solicitation direct mail context: framing valence, suggested anchorpoints, and suggested frames of reference. It discusses the importance of considering these several framing heuristics simultaneously within the same research, and offers insights on how to test these effects using experimental design concepts.

Citation:

Gerald E. Smith and Paul D. Berger (1995) ,"The Impact of Framing, Anchorpoints, and Frames of Reference on Direct Mail Charitable Contributions", in NA - Advances in Consumer Research Volume 22, eds. Frank R. Kardes and Mita Sujan, Provo, UT : Association for Consumer Research, Pages: 705-712.

Advances in Consumer Research Volume 22, 1995      Pages 705-712

THE IMPACT OF FRAMING, ANCHORPOINTS, AND FRAMES OF REFERENCE ON DIRECT MAIL CHARITABLE CONTRIBUTIONS

Gerald E. Smith, Boston College

Paul D. Berger, Boston University

ABSTRACT -

Direct mail has long been an important tool to solicit charitable contributions. While consumers have become more comfortable using direct mail, they are deluged with ever more frequent direct mail solicitations. Direct marketers also have become increasingly sophisticated, which makes it difficult for traditional charitable direct mail appeals to cut through the clutter. The purpose of this paper is to examine three framing-related creative strategies in a charitable solicitation direct mail context: framing valence, suggested anchorpoints, and suggested frames of reference. It discusses the importance of considering these several framing heuristics simultaneously within the same research, and offers insights on how to test these effects using experimental design concepts.

Potential contributors to charitable organizations have become increasingly sophisticated and discerning in their giving. American companies have restructured, downsized, and emphasized efficient allocation of expenditures. Corporate contributors focus contributions strategically on fewer causes that relate more narrowly to the company's operations or marketing (O'Hare 1991). A recent study of the 100 largest corporate donation programs in the U.S. found that more than half have developed strategic plans for charitable giving (Therrien 1992; Zetlin 1990). Slower economic growth has had a similar effect on contributions by individuals. In the U.K., for instance, the British Charities Aid Foundation reported that individual contributions in 1991 were down significantly, almost 75% from 1989 contribution levels (Hamilton 1991).

Direct mail has long been an important tool to solicit charitable contributions. It also has become an increasingly prominent direct marketing channel for products and services. In 1990, direct mail dollar sales volume exceeded $23 billion, representing 18% of advertising expenditures (Kobs 1992). This has important implications for the direct mail efforts of charitable organizations. While consumers are more comfortable using direct mail, they are deluged with ever more frequent direct mail solicitations. Kobs (1992) reports that 53 percent of all mail a consumer receives is now advertising mail, versus only 27 percent eleven years ago (page 133).

Moreover, direct marketers have become increasingly sophisticated, through industry consultants, college and university direct marketing programs, and hundreds of books on the topic. Their use of advanced creative strategies to influence buyers makes it difficult for traditional charitable direct mail appeals to cut through the clutter.

A number of creative message strategies have been studied, tested, and applied in the consumer research literature, including framing strategies. In direct marketing, however, the literature on these strategies is less well developed. It is usually normative and offers practitioner guidelines. With respect to framing, for instance, one industry leader (Stone 1992) proposes that direct marketing appeals may be oriented toward one's desire to gain: to make money, to save time, to avoid effort, to achieve comfort, to have health, to enjoy pleasure, to be clean, to be praised, to be popular. Or, they may be oriented toward one's desire to avoid loss: to avoid criticism, to keep possessions, to avoid physical pain, to avoid loss of reputation, to avoid loss of money, or to avoid trouble.

The purpose of this paper is to examine three framing-related creative strategies in a charitable solicitation direct mail context: framing valence, suggested anchorpoints, and suggested frames of reference. The paper discusses the theory behind these framing-related concepts, their application to direct marketing, and their hypothesized effects.

The paper makes several important contributions to the direct marketing and consumer research literatures. It discusses how framing-related heuristics may be applied and tested in direct marketing. It also hypothesizes how these heuristics may influence behaviors in the field of charitable contributions, a topic that has been unexplored to date. It hypothesizes how framing influences actual donor behavior, i.e., likelihood of giving and average gift, rather than attitudinal measures, such as beliefs, affect, or intent to purchase. It discusses the importance of considering these several framing heuristics simultaneously within the same research, rather than independently, and offers insights on how to test these effects using experimental design concepts.

FRAMING, ANCHORPOINTS, AND FRAMES OF REFERENCE

We distinguish among three related effects on consumer decision making with respect to charitable direct mail appeals: framing, anchorpoints, and suggested frames of reference (see Smith and Wortzel 1993). All three flow from prospect theory (Kahneman and Tversky 1979; Tversky and Kahneman 1992), or from Tversky and Kahneman's (1974) anchor and adjustment effect.

Prospect theory proposes that under uncertainty decision makers employ heuristic processing. Decision outcomes are coded categorically as gains or losses relative to a neutral reference point. Decision makers are more responsive to losses than to gains, i.e., losses loom larger than gains. Certainty is overweighted, and uncertainty is underweighted. People are risk averse with respect to gains and risk seeking with respect to losses; that is, they prefer a small certain gain to a large uncertain gain, and they prefer a large uncertain loss to a small certain loss.

Applying Kahneman and Tversky's logic to the domain of charitable contributions suggests that there are three ways to influence a decision maker's frame of reference: (1) by influencing reference point formation; (2) by describing the valence of alternatives relative to the reference point, i.e., in positive or negative terms; and (3) by changing buyers' perceptions of decision certainty. In this research we focus on the first two effects.

We refer to framing as framing valence; that is, describing product performance information or decision outcomes in either positive or negative terms (Levin and Gaeth 1988; Meyerowitz and Chaiken 1987; Maheswaran and Meyers-Levy 1990). Suggested anchorpoints refer to the presentation of a specific starting value to make decision judgments, e.g., a suggested retail price or donation level (Tversky and Kahneman 1974). Suggested frames of reference evoke comparisons of more encompassing decision frames. In this research we test two types. Experiential frames of reference are consumer focused, and compare a consumer's predecision need status to a positive or negative post-decision outcome; the consequent outcome is contingent on the decision. Product frames of reference are product focused, and compare the product performance of one product to that of another, or compare product performance to some external performance standard.

The consumer research literatures for each of these effects offer important insights that help explain and predict how framing-related effects may influence consumer behavior towards charitable contributions. We next discuss each in turn.

HYPOTHESIZED EFFECTS ON CHARITABLE DIRECT MAIL APPEALS

The Effect of Framing

Empirical framing research gives evidence that buyers are motivated by the prospect of loss aversion. Meyerowitz and Chaiken (1987) framed the positive (negative) consequences of (not) performing monthly breast self-examinations (BSE) for cancer. Stressing BSE's negative consequences was more persuasive than emphasizing its positive consequences. Maheswaran and Meyers-Levy (1990) framed message content regarding cholesterol testing in terms of "you stand to gain (lose) important health benefits if you take (fail to take) the initiative to learn what your current cholesterol count is." Negatively framed messages were more persuasive with subjects who were more involved in health and cholesterol-related issues. Positively framed messages were more persuasive with less involved subjects.

Puto's work in industrial procurement (Puto 1987; Qualls and Puto 1989; Puto, Patton, and King 1985; see also Rowe and Puto 1987) found that framing influenced reference point formation, and hence buyers' perceptions of gains and losses. Consistent with prospect theory, positively framed sales messages led to risk averse buyer behavior; negatively framed messages led to risk seeking behavior. Levin and Gaeth (1988) found that labeling a product attribute in negative terms (25% fat) led to less favorable consumer evaluations than labeling in positive terms (75% lean); the framing effect was reduced after product trial.

These findings suggest that negatively framed direct mail appeals for charitable contributions will be more persuasive than positively framed appeals. Yet, direct mail appeals for charitable contributions differ from other product or service purchase decisions studied in previous framing research. With direct mail charitable appeals actual decision behavior is observed, whereas with most other framing studies attitudinal measures have been observed. Two behavioral measures are relevant to the success of charitable direct mail appeals: the likelihood of making a donation, and the magnitude of the gift.

In most previous framing studies the decision maker benefited directly from the decision: personal health benefits (Meyerowitz and Chaiken 1987; Maheswaran and Meyers-Levy 1990; Levin and Gaeth 1988); job performance as an industrial purchase agent (Puto 1987; Qualls and Puto 1989; Puto, Patton, and King 1985). Charitable contributions result in indirect benefits to the donor. Charitable organizations reinforce these benefits by prominently publicizing the donor's gift-giving status, by offering special organizational benefits, etc.

The framing literature also frequently addresses risky purchase contexts. Charitable contributions, by contrast, involve little or no risk, unless one is motivated by possible perceptions of social risk, i.e., the risk of being viewed unfavorably by one's peers. On the other hand, many charitable appeals attempt to make the donor feel like a member of the organization such that risk and loss to the organization affects all members of the organization. Hence, donors are motivated on behalf of the organization to take action to avoid organizational risk and loss.

The shared risk that donor-members of the organization perceive is likely to be felt most acutely by those who care most about and are most involved in the organization. These donors perhaps have been associated with the organization longer, or engage in lifestyles and activities that are consistent with the organization's mission, i.e., a form of enduring involvement (Bloch and Richins 1983; Houston and Rothschild 1978). Maheswaran and Meyers-Levy (1990) give evidence that involvement moderates response to framing. Thus, the greater the perception of potential organizational risk or loss the more likely these high involvement donors will increase the magnitude of their contribution. Nonetheless, this may have little effect on the number of donors responding to a direct mail appeal, particularly if donors on the margin are less involved, or only peripherally interested in the charity's mission.

Hence, we propose that

P_1a: Negatively framed direct mail appeals for charitable contributions will yield greater average donations than positively framed appeals.

P_1b: Negatively framed direct mail appeals for charitable contributions will not yield greater likelihood of response relative to positively framed appeals.

The Effect of Suggested Anchorpoints

Tversky and Kahneman (1974) gave early evidence of the effect of anchorpoints. They noted that people make heuristic judgments relative to specific anchorpoints to reduce decision processing and mental effort. Typically, subjects bias perceptual judgments in the direction of the anchorpoint. For example, when asked to estimate the percentage of African countries in the United Nations, subjects who were given higher starting values evoked higher estimates, and subjects given lower starting values evoked lower estimates.

A number of empirical papers in marketing have demonstrated anchor and adjustment in decision making. For example, Fraser, Hite and Sauer (1988), found that the presence of a $20 "suggested contribution level" on a donation request resulted in average gifts nearly three times the average gift of a control group that received no suggested anchorpoint appeal. Brockner et. al. (1984) found that when a specific dollar amount was mentioned in telephone and face-to-face fundraising, subjects were more likely to comply and make a pledge than when no amount was mentioned. Subjects also pledged greater amounts when exposed to higher suggested anchorpoint treatments. Reingen (1982) found that exposing potential donors to a list of other compliers significantly increased the number of donors, but not the average gift. In a related experiment, the size of compliers' donations was varied, which led to changes in the magnitude of donations.

In the pricing literature, Urbany, Bearden, and Weilbaker (1988) tested different levels of "advertised reference prices." They found that even though subjects attitudinally rejected exaggerated reference prices (150 percent greater than the actual sale price), these prices still positively influenced perceptions of value and estimates of market price. Berkowitz and Walton (1980) found that deeper price discounts relative to a comparison "regular price" evoked greater perceptions of value in three separate product categories. Della Bitta, Monroe, and McGinnis (1981) found that larger differences between sale price and regular price (differences ranging from 10 percent to 50 percent) led to greater perceived savings and perceived value for money. Mobley, Bearden, and Teel (1988) also found greater perceptions of value for larger "advertised price reductions."

The Fraser, Hite, and Sauer (1988), Brockner et. al. (1984), and Reingen (1982) studies give evidence that suggested anchors influence the magnitude of gifts in the direction of the anchor. This is consistent with the pricing literature where suggested reference prices influenced buyers perceptions of attitudinal measures such as perceived value and savings. Fraser, Hite, and Sauer (1988) tested the presence or absence of a suggested anchor. Brockner et. al. (1984) tested different levels, as well as presence or absence of a suggested anchor.

The Brockner et. al. (1984) study also gives evidence that providing subjects with a concrete suggested anchor influenced the magnitude of a pledge. This is consistent with the conceptualization of suggested anchors as a decision heuristic that allows donors to quickly infer approximately what would be a reasonable gift. Their results were inconclusive, however, about whether the size of the suggested anchor influences likelihood of making a pledge. If donations were seen as a form of price given in exchange for benefits derived from a charitable organization, it seems reasonable to suggest that smaller suggested anchorpoints would yield larger levels of compliance.

Hence we propose that the magnitude of the suggested anchor will have the following effects:

P_2a: Greater suggested anchorpoints will yield greater average gifts.

P_2b: Smaller suggested anchorpoints will yield greater likelihood of response.

The Effect of Suggested Frames of Reference

Direct marketing practitioners suggest several rules of thumb that relate to suggested frames of reference: (1) Specific episode narratives outpull statistics; (2) Examples are more credible than exhortation; (3) Coupling examples with exhortation geometrically expands the emotional impact of a message (Lewis 1992).

One type of suggested frame of reference involves episodic or narrative forms of information. It provides respondents with frames of reference that are holistic, experiential, or image oriented. We call these experiential frames of reference. They involve explicit comparisons of a personal need state (before) and a consequent outcome (after), which may be framed in positive or negative terms. The outcome is contingent on the decision. In advertising, for example, many weight loss advertisements provide vivid, and frequently visual, frames of reference to potential buyers. Other outcome frames of reference in advertising present explicit points of reference relating to personal emergencies (need state before), such as a fire, hurricane, flood, accident, theft, or illness. The consequent outcome (after), according to these advertisements, depends on the consumer's decision with respect to insurance, security systems, or safety equipment. See Wiener, Gentry and Miller (1986) and Gentry, Wiener and Burnett (1987) for examples in the literature.

The following excerpt from a Chrysler advertisement illustrates:

HEADLINE: "An open letter to Lee Iacocca."

COPY: "Dear Mr. Iacocca: I sat in an ambulance today looking at my 20 year old daughter Kelly's beautiful face. She still has her beautiful face and her life because of you. . . Today a truck hit her at about 60 mph. It took them 1/2 an hour to get her out of what was left of her new car. The police said she would surely be dead if not for her airbag. . . I thanked God for you and your deep appreciation and dedication to automobile safety. . . . [signed]"

Fortunately, relative to the need state (an automobile accident), for these buyers the outcome was positive because they had made the decision to buy a Chrysler car with an airbag.

Another type of suggested frame of reference focuses on product image and product performance. It provides respondents with comparative frames of reference that are concrete, factual, and product oriented, often along specific product attributes. We call these product frames of reference. Such frames often compare one product to another, or compare a product's performance to that of an established frame of reference in the industry. The decision is framed in terms of superior performance (e.g., our brand) relative to a known frame of reference (e.g., a leading brand). Comparative advertising of a challenger brand to a leader brand illustrates a commonly used product frame of reference.

Recent empirical research by Smith and Wortzel (1993) and Smith (1992) gives evidence that the presence of suggested frames of reference in advertising favorably influences intent to purchase consumer durables (video cameras). With respect to charitable direct mail appeals, however, the only trace of evidence we have that suggested frames of reference may be effective in such appeals is embodied in the rules of thumb highlighted at the beginning of this section. While not reported in the literature but most likely studied in unreported direct marketing tests, these rules imply that narrative episodic information, similar to our experiential frames of reference, appear to have been effective in influencing donor response to charitable direct mail appeals, although it is difficult to infer the extent to which compliance, average gift, or both were affected.

We propose that experiential frames of reference will be more effective at influencing compliance than product frames of reference. The reason for this is that donors with low involvement in the organization are more likely to attend to narrative, story-oriented copy than to statistical or exhortation copy. These low involvement donors are likely to be infrequent contributors. We expect that product frames will be effective in raising the average gift, particularly for donors who are more involved in the organization and its mission, and hence already relatively frequent contributors. These donors should relate more to statistical information about the institution and its performance and status relative to other institutions.

Hence, we propose that

P_3a: The presence of an experiential frame of reference will yield greater compliance (i.e., higher response rate) than the presence of a product frame; this effect will be even greater for those who contributed less frequently in the past.

P_3b: The presence of a product frame will yield a greater average gift than the presence of an experiential frame; this effect will be even greater for those who contributed more frequently in the past.

METHODOLOGICAL ISSUES

A number of methodological issues are particularly relevant to the study of these framing-related heuristics because of the possibility of potential interaction effects. Direct mail appeals, for instance, often employ not just one framing-related heuristic, but several simultaneously. For example, a charitable organization may frame an appeal in negative terms, as a potential loss to the organization if the respondent does not comply, while also suggesting either a low or high monetary amount as an anchorpoint. Consequently it is important to study not only the individual effects of these decision heuristics, but their interaction effects as well. This has implications for the design of the experiment, as well as related issues such as experimental context and operationalization.

Selection of Experimental Context (Charitable Cause)

We refer to experimental context as the charitable cause in which to test framing-related heuristic effects. Two considerations are important. First, the researchers must be able to manipulate either positive or negative framing executions in the experimental context in such a way that these executions are viewed by respondents as credible and reasonably natural. Second, the experimental context must allow for simultaneous testing of both product frames of reference and experiential frames of reference.

From a framing valence perspective, charitable or non-profit organizations can be divided into satisfying two types of needs: those that enhance utility, and those that preserve utility (see Kahn and Meyer 1991 for a similar distinction of utility enhancing and utility preserving attributes; and Rossiter and Percy 1985, 1987 for a related discussion of transformational and informational communications).

Utility enhancing charities focus on positively valenced needs and motivations. They transform the recipient of the charity's services through sensory gratification, intellectual stimulation, or social or societal recognition or approval. Utility enhancing charities include, for example, the arts, literature, and many areas of scientific research such as space exploration or materials technologies. In such contexts positively framed appeals stress opportunities to gain; negatively framed appeals stress foregoing the opportunity to gain.

For example, college education may be framed as improving the quality of one's life by helping one fulfill a more satisfying lifestyle embodying career opportunity, income, and status. This emphasizes positive, gain-oriented, aspects of college education. College education also may be framed in terms of negative, loss-oriented dimensions. For example, a student's inability to complete a college education because of insufficient financial support may place him/her at a disadvantage relative to his/her peers. Consequently, the student struggles to get along on less income, with more career obstacles, and with fewer and less attractive opportunities.

Utility preserving charitable causes focus on negatively valenced needs and motivations. They preserve, protect, or remedy problems that affect recipients of the charity's services. These charities include, for example, research for serious diseases, such as cancer, heart, muscular dystrophy, or AIDS. They also address personal or societal needs such as hunger relief, disaster relief, environmental pollution, or resource depletion. In these contexts positively framed appeals stress the likelihood of avoiding a loss; negatively framed appeals stress the likelihood of incurring a loss.

Second, the choice of experimental context must also consider how to operationalize experiential and product frames of reference. College education, for example, offers opportunities to credibly operationalize experiential frames of reference using student experiences, e.g., a student with an acute financial need whose personal outcome is influenced by a donor's decision. The intercompetitive climate of university education also provides opportunities to operationalize product frames of reference, e.g., "our university's spending on educational programs versus other universities' spending".

Operationalizing Anchorpoints, Frames of Reference, and Framing

Operationalizing framing-related treatment manipulations involves several important considerations. These are particularly applicable in situations where the charity mails a solicitation appeal to a large number of previous donors, with a wide range of previous donations. For example, what level of donation should the charity suggest to the potential donor? For target respondents who have never previously given to the charity, a direct mail appeal may suggest a uniform suggested donation level, e.g., $100, or may suggest a range of anchorpoints in the same appeal, e.g., $25, $50, $75, $100, $ .

For target respondents who have previously given to the charity the operationalization of the suggested anchor is more complex, and more important. To suggest an anchorpoint in the current appeal that is lower than last year's donation may be misinterpreted by respondents, e.g., the organization doesn't need more money, is poorly managed, didn't appreciate my previous gift. Direct mail appeals to previous donors virtually require a customized suggested donation that is consistent with the donor's previous giving history.

Fortunately, data processing technology allows the charity manager to easily customize suggested donation appeals. However, care must be taken to ensure that this is handled appropriately. For instance, for a person who gave $210 last year, a simple linear algorithm might suggest a 15% increase this year to $241.50. While customized, to the donor such computer automated appeals are not only transparent, but seem excessively precise, artificial, and, in the end, impersonal. A more natural appeal, of course, would be to round up to a suggested donation of $250.

The issue of rounding is more complex, since rounding levels may need to differ for different levels of suggested donation. For example, for a person who gave $60 last year, a suggested 15% increase this year to $69, rounded up to the nearest $10 increment, i.e., to $70, might be reasonable. However, to a donor who gave $2,500 last year, rounding up in similar $10 increments might seem trivially precise. Here, a calculated 15% increase to $2,875 would be rounded to $2,880. It might be more reasonable to round to a suggested anchorpoint of $3,000.

A number of operational rules are possible to address rounding. We propose one that may be useful for practical implementation, and for empirically testing different anchorpoint levels in an experimental context. To illustrate, assume that a charity solicits donations from previous individual donors. Previous donations may be categorized into groups, ranging from smaller gifts (under $50) to larger gifts (over $2,500). The calculated amount for the suggested anchor then may be rounded up by an appropriate category rounding level. Table 1 shows a summary of possible donation categories and their respective rounding levels. The rounding levels were determined by the following rule: Round to a multiple of 10% of the highest donation for the category. Thus, for a donor who gave $600 last year, a calculated 15% increase this year to $690 would be rounded to a suggested anchor of $700.

Table 2 shows examples of suggested anchorpoint calculations for two potential respondents, one exposed to a suggested anchor treatment 10% higher than the last donation, the other to a treatment 50% higher than the last donation. The range of suggested anchor amounts shown in the table (i.e., +10 percent to +50 percent) is consistent with previous price anchorpoint research. For example, Della Bitta, Monroe, and McGinnis (1981) tested price discounts ranging from 10 to 50 percent. Keiser and Krum (1976) tested discounts of 50 percent. At the extreme, Urbany, Bearden, and Weilbaker (1988) tested exaggerated reference prices of 150 percent greater than the sale price.

TABLE 1

SUGGESTED ANCHORPOINT DONOR CATEGORIES AND ROUNDING LEVELS

TABLE 2

EXAMPLE OF SUGGESTED ANCHORPOINT CALCULATION FOR TWO POTENTIAL RESPONDENTS

TABLE 3

OPERATIONALIZATIONS OF EXPERIENTIAL AND PRODUCT FRAMES OF REFERENCE, WITH FRAMING VALENCE

The operationalization of frames of reference is dependent on the context of the direct mail appeal. Experiential frames of reference involve explicit comparisons of a personal need state (before) and a consequent outcome (after); the outcome is contingent on the decision. Table 3 illustrates the operationalization of an experiential frame of reference, for both positively and negatively framed conditions, for a potential educational institution's fundraising direct mail appeal. Here, potential donors are presented with an episodic narrative (in this case based on a true incident) that presents a sudden need state (before) of a student losing financial assistance, forcing her to discontinue her education. In the positively framed appeal the consequent outcome (after) is positive, i.e., she is quickly reinstated in school because the university has additional financial aid available to provide emergency assistance. In the negatively framed appeal the consequent outcome is negative, i.e., she must defer her education, because financial aid was in short supply.

Product frames of reference compare one product to another, or compare a product's performance to that of an established frame of reference. The decision is framed in terms of superior performance (e.g., our brand) relative to the frame of reference. The other half of table 3 illustrates the operationalization of a product frame of reference, for both positively and negatively framed conditions, for an educational institution. The dimension used for performance comparison is money spent per student on educational programs for students. Respondents are presented with the university's per student spending ($15,997) versus the spending of "comparable national universities." Positively framed appeals emphasize that financial support will enable the university to "continue to increase and surpass other universities." Negatively framed appeals warn that failure to support may cause the university to "fall behind other universities."

Experimental Design Issues

As noted earlier, it is important to study the interaction effects as well as the individual effects of framing valence, frames of reference, and suggested anchorpoints. One way to achieve this is to design a simple 2x2x2 design with framing valence at two levels (positive, negative), suggested anchorpoint at two levels (high, low), and frame of reference at two levels (product frame, experiential frame). In addition to simplicity, this 8 cell design offers the ability to test all main effects, all 2-way interactions, and the 3-way interaction among the three major framing-related heuristics.

However, this design precludes the possibility of testing the potential interaction effect between presence/absence of product frame of reference and presence/absence of experiential frame of reference, or of testing the interaction effect of, say, presence/absence of experiential frame and level of framing valence, or presence/absence of product frame and level of framing valence. Moreover, charities employing direct mail may want to employ both product frame of reference and experiential frame of reference in the same appeal.

This requires experimentally treating presence/absence of product frame of reference as one factor and presence/absence of experiential frame of reference as another factor. In this case a two-way interaction between factors "Product Frame" and "Framing Valence" can be described by saying that "the impact of the presence of a product frame differs depending on the level of framing valence," or "the impact of framing valence differs depending on the presence/absence of a Product Frame." Both descriptions are numerically equivalent; one may be more contextually useful.

Treating product frame and experiential frame as two separate factors, each with two levels (present, absent), yields a four-factor full factorial design as follows:

TABLE

This design allows testing the interaction of product frame of reference and experiential frame of reference, BC. It also allows testing the remaining five 2-way interactions among the other factors, i.e., AB, AC, AD, BD, and CD, as well as the four 3-way interactions, i.e., ABC, ABD, BCD, ACD, and the 4-way interaction, ABCD.

A potential disadvantage of this design to charities using direct mail is the number of treatment cells required for implementation, 16. This may be unwieldy and costly. The experimental design literature offers several possible solutions. One is to utilize a fractional factorial design, rather than a full factorial design. For example, a 2^4-1 fractional factorial design with defining relation/contrast I=ABCD results in each of the 4 main effects being aliased with a 3-factor interaction, and each of the 6 two factor interactions being aliased with another two factor interaction, as follows:

TABLE

See Table 4 for detailed specification of cells and treatment levels for this design. For details of alias patterns in fractional factorial designs, see for example Davies (1984). Given the typical assumption that all 3-way and higher order interactions are zero (or negligible), the main effects thus may be cleanly (and orthogonally) estimated. The 2-way interactions, however, are aliased in pairs, which means that it is necessary to provide independent evidence, from other studies, or separate tests, that some of the two way interactions are in fact zero.

Previous framing research in an advertising context (Smith and Berger 1994) gives evidence of significant two-way interaction effects between product frames of reference and framing valence (BD), and between experiential frames of reference and framing valence (CD), but of non-significant two-way interaction effects between suggested anchorpoint and product frame of reference (AB), between suggested anchorpoint and experiential frame of reference (AC), and between product frame of reference and experiential frame of reference (BC). If it were reasonable to assume that these patterns held across experimental context, then the AB, AC, and BC 2-factor interaction effects could be assumed to be zero (or negligible). This would allow clean and orthogonal estimation of the remaining 2-factor interactions: AD, BD, and CD.

CONCLUSION

Direct mail has long been a staple fundraising technique for charitable organizations. In recent years, however, buyers have become more deluged with mail advertising appeals, while direct mail marketers have become increasingly sophisticated. This paper conceptualizes, hypothesizes, and describes how to test the effect of three creative strategies that should be useful to charities' direct mail efforts. The paper is part of a larger project designed to empirically test these effects.

Based on the consumer behavior literature we distinguished between framing, describing product performance or decision outcomes in positive or negative terms; suggested anchorpoints, the presentation of a specific starting value to make decision judgments; and suggested frames of reference, the comparison of post-decision outcomes with a pre-decision need state for experiential frames, and the presentation of our product's performance versus other performance frames of reference for product frames.

TABLE 4

EXPERIMENTAL DESIGN

This work complements related efforts in advertising and marketing communications, and hopefully will help stimulate further research in the fields of direct marketing and framing in general.

REFERENCES

Berkowitz, Eric N. and John R. Walton (1980), "Contextual Influences on Consumer Price Responses: An Experimental Analysis," Journal of Marketing Research, Vol 17 (August), 349-358.

Bloch, Peter H. and Marsha L. Richins (1983), "A Theoretical Model for the Study of Product Importance Perceptions," Journal of Marketing 47 (Summer), 69-81.

Brockner, Joel et. al. (1984), "Organizational Fundraising: Further Evidence on the Effect of Legitimizing Small Donations," Journal of Consumer Research 11 (June), 611-614.

Davies, O. L. (1984). The Design and Analysis of Industrial Experiments. New York: Hafner, Third Edition.

Della Bitta, Albert J., Kent B. Monroe, and John M. McGinnis (1981), "Consumer Perceptions of Comparative Price Advertisements," Journal of Marketing Research, Vol. 18 (November), 416-427.

Fraser, Cynthia, Robert E. Hite and Paul L. Sauer (1988), "Increasing Contributions in Solicitation Campaigns: The Use of Large and Small Anchorpoints," Journal of Consumer Research, 15 (September), 284-287.

Gentry, James W., Joshua L. Wiener, and Melissa Burnett (1987), "The Story, the Frame, and the Choice," in Melanie Wallendorf and Paul Anderson, eds., Advances in Consumer Research, Vol. 14. Provo, UT: Association for Consumer Research, 198-202.

Hamilton, Sally (1991), "Cashing In on Good Works," Business-London (July), 99-101.

Houston, Michael J. and Michael L. Rothschild (1978), "Conceptual and Methodological Perspectives in Involvement," in Research Frontiers in Marketing: Dialogues and Directions, S. Jain, ed. Chicago: American Marketing Association, 184-7.

Kahn, Barbara E., and Robert J. Meyer (1991), "Consumer Multiattribute Judgments under Attribute-Weight Uncertainty," Journal of Consumer Research, 17 (March), 508-522.

Kahn, Barbara E., and Robert J. Meyer (1991), "Consumer Multiattribute Judgments under Attribute-Weight Uncertainty," Journal of Consumer Research, 17 (March), 508-522.

Kahneman, Daniel and Amos Tversky (1979), "Prospect Theory: An Analysis of Decision Under Risk," Econometrica, 47 (March), 263-291.

Keiser, Stephen K. and James R. Krum (1976), "Consumer Perceptions of Retail Advertising With Overstated Price Savings," Journal of Retailing, Vol. 52 (Fall), 27-36.

Kobs, Jim (1992). Profitable Direct Marketing, second edition. Lincolnwood, IL: NTC Books.

Levin, Irwin and Gary J. Gaeth (1988), "How Consumers Are Affected by the Framing of Attribute Information Before and After Consuming the Product," Journal of Consumer Research, 15 (December), 374-378.

Lewis, Herschell Gordon (1992), "Fifty of the Easiest Ways to Begin An Effective Sales Letter," Direct Marketing (October), 16-19.

Maheswaran, Durairaj and Joan Meyers-Levy (1990), "The Influence of Message Framing and Issue Involvement," Journal of Marketing Research, 27 (August), 361-367. -

Meyerowitz, Beth E. and Shelly Chaiken (1987), "The Effect of Message Framing on Breast Self-Examination Attitudes, Intentions, and Behavior," Journal of Personality and Social Psychology, Vol. 52, No. 3, 500-510.

Mobley, Mary F., William O. Bearden, and Jesse E. Teel (1988), "An Investigation of Individual Responses to Tensile Price Claims," Journal of Consumer Research, 15 (September), 273-279.

O' Hare, Barbara Clark (1991), "Good Deeds Are Good Business," American Demographics 13 (September), 38-42.

Puto, Christopher P. (1987), "The Framing of Buying Decisions," Journal of Consumer Research, Vol. 14 (December), 301-315.

Puto, Christopher P., Wesley E. Patton III, and Ronald H. King (1985), "Risk Handling Strategies in Industrial Vendor Selection Decisions," Journal of Marketing 49, 89-98.

Qualls, William J. and Christopher P. Puto (1989), "Organizational Climate and Decision Framing: An Integrated Approach to Analyzing Industrial Buying Decisions," Journal of Marketing Research, Vol. 26 (May), 179-192.

Reingen, Peter H. (1982), "Test of a List Procedure for Inducing Compliance with a Request to Donate Money," Journal of Applied Psychology 67 (February), 110-118.

Rowe, Debra and Christopher P. Puto (1987), "Do Consumers' Reference Points Affect Their Buying Decisions?" in Melanie Wallendorf and Paul Anderson, eds., Advances in Consumer Research, Vol. 14. Provo, UT: Association for Consumer Research.

Smith, Gerald E. (1992), "Prior Knowledge and the Effectiveness of Message Frames in Advertising," Unpublished dissertation, Boston University.

Smith, Gerald E. and Paul D. Berger (1994), "The Impact of Anchor Framing and Prior Knowledge on Advertising Effectiveness," Working Paper, Boston College.

Smith, Gerald E., and Lawrence H. Wortzel (1993). Prior Knowledge and the Effect of Message Frames in Advertising. Cambridge, MA: Marketing Science Institute.

Stone, Bob (1992). Successful Direct Marketing Methods, Fourth Edition. Lincolnwood, IL: NTC Business Books.

Therrien, Lois (1991), "Corporate Generosity is Greatly Depreciated," Business Week (November 2), 118-120.

Tversky, Amos and Daniel Kahneman (1974), "Judgment Under Uncertainty: Heuristics and Biases," Science 185, 1124-31.

Tversky, Amos and Daniel Kahneman (1992), "Advances in Prospect Theory: Cumulative Representation of Uncertainty," Journal of Risk and Uncertainty.

Urbany, Joel E., William O. Bearden, and Dan C. Weilbaker (1988), "The Effect of Plausible and Exaggerated Reference Prices on Consumer Perceptions and Price Search," Journal of Consumer Research, Vol. 15 (June), 95-110.

Wiener, Joshua L., James W. Gentry, and Ronald K. Miller (1986), "The Framing of the Insurance Purchase Decision," in Richard J. Lutz, ed., Advances in Consumer Research, Vol. 13. Provo, UT: Association for Consumer Research.

Zetlin, Minda (1990), "Companies Find Profit in Corporate Giving," Management Review 79 (December), 10-15.

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