Framing Dynamics: Measurement Issues and Perspectives

Donald J. Hempel, University of Connecticut
Harold Z. Daniel, University of Connecticut
ABSTRACT - Consumer researchers have historically represented decision frames as a manipulated variable within an experimental design for observing reactions to differing reference points. The typical research designs impose models that represent consumer decision frames as stable over time. A different conceptualization is presented here to focus on actual market conditions in which consumers develop and react to their own frames of reference, and operate with multiple decision frames that evolve over time. This dynamic framework provides a means of linking consumer assessments of value and risk through reference prices.
[ to cite ]:
Donald J. Hempel and Harold Z. Daniel (1993) ,"Framing Dynamics: Measurement Issues and Perspectives", in NA - Advances in Consumer Research Volume 20, eds. Leigh McAlister and Michael L. Rothschild, Provo, UT : Association for Consumer Research, Pages: 273-279.

Advances in Consumer Research Volume 20, 1993      Pages 273-279

FRAMING DYNAMICS: MEASUREMENT ISSUES AND PERSPECTIVES

Donald J. Hempel, University of Connecticut

Harold Z. Daniel, University of Connecticut

ABSTRACT -

Consumer researchers have historically represented decision frames as a manipulated variable within an experimental design for observing reactions to differing reference points. The typical research designs impose models that represent consumer decision frames as stable over time. A different conceptualization is presented here to focus on actual market conditions in which consumers develop and react to their own frames of reference, and operate with multiple decision frames that evolve over time. This dynamic framework provides a means of linking consumer assessments of value and risk through reference prices.

Research on consumer decision-making has drawn heavily from expected utility models with their assumption that consistent mental structures are employed in the evaluation of product alternatives. Prospect theory (Kahneman and Tversky 1979) conceptualizes evaluation as influenced by reference points in the assessment context or framework. Cognitive theorists (Rosch and Mervis 1975) also argue that reference points serve as "benchmarks" or "anchors" for comparisons. When internalized by individuals, these scales and contrasts are fundamental influences on their perceptions, information processing and judgments. Applications of prospect theory to consumer behavior suggest that product alternatives are presented (coded) relative to reference points against which they are judged as a gain or loss. Little is known about how decision frames are shaped and changed, or how these dynamics influence consumer perceptions of value (Elliot and Archibald 1989).

Issues and conceptualizations concerning the value assessment process have been summarized by Zeithaml (1988). The importance of framing in understanding relationships between value and risk is highlighted by Tversky and Kahneman (1986). Strategic marketing concerns for integrating these perspectives require more information about the evolutionary nature of framing effects (Kalwani, Yim, Rinne and Sugita 1990). The purpose of this paper is to identify some of the conceptualization and measurement issues associated with linking value and risk constructs to the evolution of decision frames we call framing dynamics.

THEORETICAL PERSPECTIVES

Value perception is a function of how individuals frame buying decisions (Puto 1987, Zeithaml 1988, Qualls and Puto 1989). Zeithaml (1988) suggests that the cues which signal quality and value change over time are induced by the dynamics of competition, promotional efforts, consumer tastes and available information. Such observations support our conviction that the reference points and framing context are dynamic for purchase decisions. The concept of an evolving framework of reference points stimulate concerns for both the structure of the frames imposed by decision makers as well as the dynamics of framing effects over time. Useful insights can be generated from several theoretical perspectives pertaining to the structure and evolution of decision frames and the associated effects on value-risk relationships.

Risk Perception

Theories that emphasize risk-adjusted value perception associated with decisions include Prospect Theory (Tversky and Kahneman 1986) and Venture Theory (Hogarth and Einhorn 1990). Both theories regard risk as the uncertainty in outcomes associated with a particular decision situation and postulate that an individual's evaluation of alternatives depends upon the nature of the decision frame (whether the expected outcome is viewed as a potential gain or a potential loss). Although the decision frame is regarded in both theories as the context within which a decision is made, neither theory specifically addresses the notion of evolving frame structures. The research designs typically employed draw inferences from experimental situations in which individual choices are manipulated through the frames imposed by the researcher. One could argue that such manipulations were of information rather than decision frames, leaving subjects free to formulate their own decision frames based on the manipulated information. However, these research designs are clearly not bias free. Subjects are usually placed in a problem solving situation where the artificial environment of the laboratory intensifies sensitivity to the researchers' instructions, thereby creating biased decision frames compared to the frames that subjects might form on their own. Thus, generalizations concerning framing effects derived from this research stream tend to rely on two assumptions: (1) that individuals would behave similarly when forming their own decision frames without intervention; and (2) that these decision frames are stable over time. The first assumption has been challenged (e.g., Elliott and Archibald 1989) and the latter assumption has yet to be empirically verified. Consequently, the framework derived from such studies is inadequate by itself for a complete understanding of risk-value relationships in consumer decision processes.

Value Perception

Among the theories that emphasize issues regarding value perception are Lancaster's Utility Theory (1971) and Westbrook and Reilly's Value Percept Disparity Theory (1983). While Prospect and Venture Theories regard products holistically, these theories deal with the products by their component attributes. They view products as bundles of attributes of which each possesses some intrinsic value to the consumer in terms of providing need satisfaction or personal goals and values. If personal goals and values of the individual determine the structure of decision frames, frames must evolve over time since personal values change as the individual matures.

The role of expectations in the disconfirmation paradigm of value assessment is similar to the role of reference points under Prospect and Venture theories, yielding outcomes viewed as gains or losses. For example, Thaler (1985) defines a consumer value function for differences relative to reference points rather than absolute levels. This suggests that consumer perceptions of both value and risk may be influenced by reference points in an evolving context. The dynamic role of reference points in value assessment lead to consideration of price as a framing measure (Monroe and Chapman 1987). Whereas the expectancy-disconfirmation paradigm may imply stable expectations of a product, the implications of an evolving assessment framework merit further exploration.

Reference Pricing

The notion of changing reference points is well represented in the theoretical foundations of reference price research (Winer 1988). The body of research on reference pricing is based on Assimilation-Contrast (Sherif and Hovland 1961) and Adaption Level Theories (Helson 1964). Both theories suggest that consumers determine value on the basis of reference points, such as the prices derived from current market pricing or from past product pricing (Lichtenstein and Bearden 1989). Adaptation Level and Assimilation-Contrast theories imply that shifts in the internalized reference price or latitude of acceptance can result from changes in the overall level of market prices or by gradual changes in the price of a specific brand. It can be hypothesized from this that the decision frame represented by the reference price would also be subject to change and thereby produce dynamic framing effects.

The concept of risk in decision framing can be linked to reference pricing through studies of expectation-disconfirmation theory. Della Bitta, Monroe and McGinnis (1981) examined the impact of different-sized price discrepancies from a reference point in the form of advertised discounts from a "regular" price. They found that larger discounts increased perceived value and interest while diminishing intent to search. In this situation, the adjustment of reference prices can be interpreted as risk-reduction effects that change the value assessment context or decision frame used in the choice process.

MEASUREMENT ISSUES

A basic thesis of framing dynamics is that the sensitivity of choice tasks to context effects in major purchase decisions can be represented through the linkage of featural structure (e.g., product attributes) to value structure (e.g., reference prices and perceived risk). Prior research has outlined the conceptual issues and analytical framework involved in this linkage (Hempel and Daniel, 1992). This paper concentrates on the definition of the dimensional structure of decision frames and the effect of evolving frame structures on perceived value and risk. Belk (1975) provides relevant perspectives on similar context definition problems in his conceptualization of situational variables. He distinguishes temporal perspectives from task definition in developing a dimensional taxonomy, and highlights the significance of subjective vs. objective measurements in assessments of situational effects.

Evolving sets of multiple reference points present significant problems for representing framing effects in choice models. These problems are minimized when the relevant context is interpreted as a single decision frame, such as in experimental settings where a frame is imposed and exposure is controlled. The decision situations encountered by consumers in actual market situations may require different perceptual representations because they are separated in time. Information derived from one time-related set of reference points and its structural representation (e.g., whether alternatives are perceived as gain or loss) may be altered by temporally separated sets of subsequent reference points. For example, in housing markets pre-search encounters with media reports on overall housing sales may indicate that a loss will result from sale of the buyer's present residence, while subsequent broker communications emphasizing the reduced purchase price of an alternative property may essentially reframe the transaction as a gain. These shifting context effects may account for some of the breakdowns between geometric and algebraic representations of consumer similarity judgments noted by Glazer and Nakamoto (1991).

The structural dynamics reflected in the evolving, interactive quality of the successive sets of reference points invoked by home buyers is the main concern of this paper. Operational definitions of framing dynamics require representation of both intra-frame and inter-frame relationships. Intra-frame measures are conceptualized as structural concerns based on the measurements used in prior studies of framing effects, with extensions to include frame boundary issues. Inter-frame measures are presented as dynamic concerns that focus on the differences across measurement sets and related patterns of convergence.

Framing Structure

Measurement of framing structure requires concern for both the geometric and algebraic relationships represented in consumer similarity judgments. Glazer and Nakamoto (1991) advise that it is meaningful to think in terms of both the amount and type of structure captured by alternative representations. They caution that the structure imposed by a model operates as a constraint on the empirical relations captured. The measurement considerations outlined below focus on algebraic representations of relationships among reference prices. Concerns for the related geometric or spatial relations involved are addressed in the following discussion of zones, gradients and convergence patterns.

Reference Points are the fundamental measures used to define the dimensional structure of each decision frame. As considered earlier, the theoretical basis for this interpretation is derived from Adaptation Level Theory and Assimilation-Contrast Theory. The reference points used in this paper to define value structure are the prices of housing alternatives considered within a decision frame. Assimilation-Contrast theory postulates that these intra-frame reference points will have carry-over effects (response latencies) that influence reactions to the alternatives considered in subsequent decision frames. These inter-frame relationships are considered in the following section. The actual distribution of reference points within each frame provides the basis for defining four aspects of dimensional structure for each frame:

Density - the number of alternatives considered within a specific decision frame may range from a sparse set (e.g., only high-low reference points) to a very dense set (e.g., all the alternatives considered during search).

Range - the upper and lower values of the reference points is defined by the range of prices considered within each frame. Nwokoye (1975) presents evidence that buyers use extreme prices as anchors, i.e. the highest and lowest prices evaluated.

Stability - the relative dispersion (concentration) of reference points is represented as intra-frame variance.

Centroid - the central tendency of the reference points in a decision frame is represented as a mean or median reference price. Emery (1970) suggests that the anchor price will be an average of the prices considered by the buyer.

Decision Frames are the time-related representations of buying process stages (e.g., pre-search vs. post-search) defined in terms of the buyer's information processing and decision-making behavior. For this preliminary analysis, the sequential sets of reference points considered were limited to the six decision frames (DFs) described below. Table 1 summarizes the empirical basis for operationally defining these frame boundaries.

DF1: the set of reference points defined by the present value of the purchase price of the home purchased immediately prior to the current home, the purchase offers received for this prior home, and its selling price. If the prior residence was a rental unit, the capitalized value of the monthly rental payment is substituted as the dominant reference point.

DF2: the initial reference points invoked in the problem recognition phase of the buying process. These internalized references are derived from prior purchase experiences, pre-search media exposure, and reflective knowledge activated by initial need recognition.

DF3: the reference points invoked immediately prior to beginning active external search behavior. This pre-search frame is active at the point when a decision has been made to purchase a new home, but before any homes have been visited or any real estate agent has been contacted.

DF4: the set of reference points derived from active search behavior based on the asking prices of the entire distribution of housing alternatives visited and/or considered for purchase.

DF5: the choice or post-search reference points comprised by the set of housing alternatives the buyer seriously considered purchasing.

DF6: the reference point represented by the actual purchase price. This is both an outcome of the decision process and a decision frame for evaluating purchases of related goods (e.g., structural improvements, appliances and furniture).

Each decision frame involves a frequency distribution of reference points that can be represented by various descriptive statistics. The summary statistics used to define the structural measurements include density, range, variance and central tendency. All of these framing structure measures were standardized by dividing the reference prices reported by the final purchase price. These standardized measures reduce the magnitude of reported differences across housing price levels and thereby provide a more exact basis for comparisons across frames and cases. The defined decision frames can be interpreted as fuzzy sets with significant overlap among reference points, particularly between adjacent frames. However, the added temporal dimension derived from the time (sequence) linked decision process references serve to distinguish the frame boundaries. In fact, understanding the interdependencies among the respective frames is a focus of this research.

Framing Dynamics

Measurement of framing dynamics requires concern for the evolution of structural dimensions across stages of the buying process. In essence, the concepts of reference points and decision frames must be extended to (1) examine the relationships among successive sets of reference points; and (2) interpret the effects of evolving decision frames on the relevant context for home buying decisions. These extensions dealing with the evolution of frames are operationally defined below in terms of zones, gradients and convergence rates. Table 2 provides a summary of these inter-frame measures.

Evolving sets of interacting reference points present measurement problems that are similar to those of repeated measurement designs. These basic problems stem from concerns for the interrelationships within a series of observations, including: (1) Carry-over effects - exposure to reference points in the next decision frame occurs before the effects of the prior frame reference points have "worn off"; (2) Latency effects - exposure to reference points in the next decision frame "activate" or interact with those of the previous frame; and (3) Learning effects - exposure to reference points in successive frames "limits" or converges the selection and interpretation of reference points in future frames. These concerns are usually treated as problems or difficulties to be removed by statistical controls in repeated measurement designs. In this application, their presence is interpreted as evidence of inter-frame relationships to be highlighted as measures of framing dynamics.

In non-experimental settings, measurement of inter-frame relationships is limited by the models available for separating perceptual effects from preference effects in the evolving context. For example, the range of prices buyers actually use to evaluate alternatives are determined by perceptions and judgments within boundaries constrained by initial preferences. Subsequent preferences are influenced by the reference points emerging from alternatives to which the buyer is exposed during the buying process. The resulting pattern of framing effects must be interpreted with concern for the underlying elements of reference point continuities that extend across frames. Relevant frameworks for tracking this underlying continuity are evident in studies of the process by which internal price standards are formed, particularly those that highlight the interrelationships of prices consumers perceived market prices and latitudes of price acceptance (e.g., Lichtenstein and Bearden 1989; Della Bitta, Monroe and McGinnis 1981).

Reference point definitions focus on three related aspects of framing dynamics: the range of points judged to be relevant (zones); relations among the series of points considered (gradients); and the contrast and assimilation effects represented in the internalization of reference points (patterns). Concerns for market efficiency suggest that the rate at which these differences are reconciled or converge is an important basis for assessing performance. These measurement issues and operational definitions are summarized in Table 2. A graphical summary of the relationships among these geometric measures (zones and gradients) is presented in Figure 1.

Zones: Decision-maker uncertainty is reflected in the band or range of reference points perceived to be acceptable (Stoetzel 1970). How these internal judgment scales change across time is an essential concern in research on context effects. Reference Zones represent the scale of reference points linking successive decision frames defined by the high-low range of reference points in the preceding frame. Continuity Zones represent the consistency of references points across decision frames defined by the range of one standard deviation around the price gradient based on a global (across frames) calculation of variance.

Gradients: Measures of adaption levels have been used to represent the pooled effects of the reference points used across time (Della Bitta and Monroe 1973). Issues regarding the appropriate weighting and anchoring of these changing reference regions lead to a concern for underlying commonalities that are anchored by multivariate reference points. The Price Gradient is conceptualized as the "flight path" of the expected framing dynamics. It is defined by the functional relationship between the reference points in the initial (DF1) and final (DF6) decision frames. This gradient represents the evolving standard of internal judgments by which perceived value is adjusted to the quality and price relationships encountered.

TABLE 1

OPERATIONAL DEFINITIONS OF FRAMING STRUCTURE

TABLE 2

OPERATIONAL DEFINITIONS OF FRAMING DYNAMICS

The Risk Gradient is a baseline reference extending across decision frames as an evolving measure of sacrifice and acceptable risk defined by the ratio of monthly housing payments to household income after taxes. This ratio is calculated to represent the change in income commitments (sacrifice) associated with the purchase decision and normalized to reflect the proportion of income spent on housing. It contrasts the level of risk (sacrifice) in the entry frame (DF1) relative to the risk level represented in the purchase frame (DF6). This measure is adjusted for first-time home buyers to consider prior rental payments as well as mortgage payments.

Pattern: The covariance of reference points across frames provides insights into the stability of internalized standards. The geometry of these changing context relationships may address measurement issues concerning the symmetry of psychological distances and their representation by algebraic measures. The Pattern of variance for each frame is defined relative to the variance across all frames. It is measured by a ratio of the coefficient of variance across all frames (calculated as the global mean divided by the global standard deviation), relative to the coefficient of variance for each frame (calculated as the frame mean divided by the frame range).

FIGURE 1

FRAMING DYNAMICS: ZONES AND GRADIENTS

Convergence: Price convergence represents the trend of inter-frame reference point differences relative to the price gradient extending across decision frames. It is calculated as the difference between the average (mean or median) of the current frame and the midpoint (high minus low divided by two) of reference points in the prior frame. In effect, the midpoint of each successive frame represents the preferred reference scale while the highest and lowest reference points in the previous frame are the anchors against which it is judged (see Della Bitta and Monroe 1973 for discussion of anchoring issues). Risk convergence represents the trend of reference points across frames defined relative to the risk gradient. It is calculated as an entropy measure representing the information content of each frame as a predictor of the overall risk gradient defined by the housing payments ratio.

Propositions

Empirical tests are needed to confirm framing dynamics as a valid conceptualization of choice behavior. The following propositions will be tested.

Separability: In order to conclude that framing dynamics is a more appropriate representation of the consumers' decision processes than static framing, a series of evolving frames must be distinguished from a single frame conceptualization. Therefore,

P1: The predictability of the final purchase price will be greater when both inter-frame and intra-frame variances in reference prices are represented in the predictive equation.

Latency: If the evolving sets of reference points involved in framing dynamics are to be of value in predicting consumer behavior, the distribution of points in each decision frame should contribute to the prediction of succeeding frames. That is,

P2: The interaction of reference points across decision frames will be reflected in a latency effect.

Independence: In order to validate the multiple decision frame conceptualization, each decision frame must contribute significantly to the prediction of final purchase price. Although latency effects are expected in measures of the average effects over all frames, each decision frame is expected to have a unique effect beyond this latency effect.

P3: Each decision frame will evidence a unique (contrast) effect as well as an interaction (latency) effect.

Convergence: If learning effects drive the evolution of reference points across decision frames and enhance the effectiveness of consumer decision making, intra-frame variance should decline in successive stages of the decision process.

P4: Intra-frame variance will decline across each succeeding decision frame.

Price Orientation: The informational value of price is a function of the range of reference points in relation to price and risk gradients. As a result of learning effects and the internalization of pricing standards, the meaning of price evolves from a symbolic role in communicating quality to an economic role as an indicator of sacrifice. This change in roles results in an increase in the predictability of subsequent frames.

P5a: The price gradient will be the operative orientation during the early stages of the decision process while the intra-frame variance in reference points is increasing relative to the inter-frame variance.

P5b: The risk gradient will be the operative orientation during the later stages of the decision process while the intra-frame variance in reference points is declining relative to the inter-frame variance.

P5c: The shift from a price orientation to a risk orientation will be evident in a decline of the inter-frame entropy measure.

EMPIRICAL PERSPECTIVES: HOME BUYER BEHAVIOR

Data obtained from recent home buyers provide a basis for empirically testing these relationships in a non-experimental setting. The home buying process was selected as an empirical metaphor for purchase decision settings where framing dynamics are likely to be evident. Consumers who purchase a house have been involved in the evaluation of a durable good with significant economic and social consequences. Housing purchases usually invoke high-involvement, extended problem solving behavior with multiple reference points and significant price effects. In this situation, propositions concerning the evolution of decision frames can be tested as observable and meaningful representations of market realities, and externally validated through macro measures. Housing choice behavior can be interpreted as an anchor point for theory development because it represents one extreme of an economic value continuum. Thus, empirical evidence of framing dynamics in this field setting provides useful benchmarks in assessing generalization of the model to other product classes.

TABLE 3

FRAMING DYNAMICS: FRAMS, GRADIENTS AND ZONES

Household level data were collected from 230 home buyers in a major regional housing market in 1988, prior to the economic slowdown in the region. The sampling frame consisted of three metropolitan areas, including 18 communities with a wide range of house prices and local market size differences. Survey methods were employed to obtain data from a systematic random sample of all households listed in deed transfers recorded during the year beginning April 1, 1988. In-depth personal interviews with household heads were conducted from June through October of 1989. Separate mail questionnaires with different questions for the male and female household heads were distributed following the in-home interviews. A series of telephone and mail follow ups were employed to achieve mail responses from approximately 70 percent of the households interviewed.

This data base provides an extensive and internally consistent empirical context for modeling framing dynamics and its effects on consumer choice behavior. It includes measures of prior home buying experiences, perceived knowledge and confidence levels, external information sources used, and the relative importance of product attributes. Subsequent analysis will consider these variables as moderators and complementary framing measures that may help to refine the definition of framing dynamics.

Generalizations about framing effects based on survey data are limited by the ability of respondents to recall their reference points and their willingness to accurately disclose them. The relatively high levels of product importance, buyer involvement and cognitive processing evident in home buying behavior help to minimize these limitations. Nonetheless, the process of eliciting subjective frames and the interpretation of how individuals independently impose frames on their choice is influenced by the data collection methodology used.

Methods and Preliminary Results

The space constraints imposed on this paper preclude the normal progression of this discussion through methodology, results and discussion sections. However, a preview of the analytical procedures and some preliminary results may facilitate assessment of the conceptualization and help to improve the research plan.

Repeated measurement designs (MANOVA) provide a framework for analyzing effects that are measured on several occasions for each respondent. These procedures incorporate allowances for dependencies that facilitate interpretation of differences across levels (or stages) of a factor. Linear combinations of differences (contrasts) can be specified to compare each level of a factor to the average of the levels that precede it. Thus, frames can be interpreted as levels or stages of a framing factor with concern for latency, carry-over and learning effects. These concerns are usually treated as problems or difficulties to be removed in repeated measurement designs. In this application, their presence is interpreted as evidence of interframe relationships to be highlighted as measures of framing dynamics.

Table 3 provides an empirical example of how the decision frame means and gradients differ across respondents. These results indicate that interpretation of all reference points as a single frame may result in significant loss of information about framing effects on choice behavior. The change in mean reference points from DF1 to DF2 and from DF4 to DF5 suggest that price may play different roles as the decision process evolves. The mean variance measured by intra-frame standard deviations was notably stable across frames, while the range changed significantly in DF3 and DF5 respectively. Relationships among frames reflected in the price and risk convergence measures indicates that exposure to market realities during the period of active search (DF4) may result in some significant restructuring of the perceptual context for home buying decisions.

CONCLUSIONS

This paper presents some measurement issues associated with linking framing dynamics to the interpretation of consumer value and risk perceptions. Our conceptualization of subjective framing effects adds several perspectives to understanding the value-risk relationship. First, decision frames are interpreted as an evolving multi-stage framework defining the perceptual domain in which value and risk are interpreted. Second, these successive frameworks are reformulated as dynamic sets of relevant reference points invoked by the consumer during the choice process. Finally, concerns for this evolutionary process extend existing paradigms of framing effects to include measures of change rates and convergence patterns.

This design differs from typical studies of perceptual effects in two critical areas. First, it examines framing effects in behavioral environments where consumers establish their own decision frames. The concept of framing dynamics is likely to have empirical validity in realistic market environments. It may not be evident in experimental settings where the researcher's instructions control the framing process. Second, our conceptualization of framing dynamics requires measurement of reference point relationships that extend across time-linked perceptual spaces. This adaptive environment complicates the analysis by requiring measures of state change effects. It also affords the opportunity to examine the mechanisms through which perceptions of value, risk and quality are reconciled.

REFERENCES

Belk, Russell W. (1975), "Situational Variables and Consumer Behavior," Journal of Consumer Research, 2, 157-164.

Della Bitta, A., K. Monroe and J. McGinnis (1981), "Consumer Perceptions of Comparative Price Advertisements," Journal of Marketing Research, 18, 419-27.

Della Bitta, Albert J. and Kent Monroe (1974), "The Influence of Adaptation Levels on Subjective Price Perceptions," Advances in Consumer Research, 1, 359-69.

Elliott, Catherine S. and Robert B. Archibald (1989), "Subjective Framing and Attitudes Towards Risk," Journal of Economic Psychology, 10, 321-328.

Emery, Fred (1970), "Some Psychological Aspects of Price," in Pricing Strategy, B. Taylor and G. Wills, eds. Princeton, NJ: Brandon System Press, 98-111.

Glazer, Rashi and Kent Nakamoto (1991) "Cognitive Geometry: An Analysis of Structure Underlying Representations of Similarity," Marketing Science, 10, 3, 205-228.

Helson, H. (1964) Adaptation-Level Theory, New York: Harper and Rowe.

Hempel, Donald J. and Harold Z. Daniel (1992), "Framing Dynamics: Linking Risk and Value," in European Marketing Academy Proceedings, K. Grunert and D. Fuglede, eds., 599-617.

Hogarth, Robin M. and Hillel J. Einhorn (1990), "Venture Theory: A Model of Decision Weights," Management Science, 36, 7, 780-803.

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

Kalwani, Manohar, Chi Yim, Heikki Rinne and Yoshi Sugita (1990), "A Price Expectations Model of Customer Brand Choice," Journal of Marketing Research, 27, 251-62.

Lancaster, K. J., 1971. Consumer Demand: A New Approach, Columbia University Press, New York.

Lichtenstein, Donald R. and William O. Bearden (1989), "Contextual Influences on

Perceptions of Merchant-Supplied Reference Prices," Journal of Consumer Research, 16, 55-66.

Monroe, Kent B. and Joseph D. Chapman, 1987. Framing Effects on Buyers' Subjective Product Evaluations. Advances in Consumer Research. 14, 193-197.

Nwokoye, Nonyelu G. (1975), "Subjective Judgements of Price: The Effects of Price Parameters on Adaptation Levels," Proceedings of Fall Conference. Chicago: American Marketing Association.

Qualls, William J. and Christopher Puto (1989), "Organizational Climate and Decision Framing: An Integrated Approach to Analyzing Industrial Buying Decisions," Journal of Marketing Research, 26, 179-92.

Puto, Christopher P. (1987), "The Framing of Buying Decisions," Journal of Consumer Research, 14, 301-314.

Rosch, Eleanor and Carolyn B. Mervis (1975), "Family Resemblances: Studies in the Internal Structure of Categories," Cognitive Psychology, 7, 573-605.

Sherif, M. and C. Hovland (1961), Social Judgement, New Haven, Connecticut: Yale University Press.

Stoetzel, J. (1970), "Psychological/Sociological Aspects of Price," in Pricing Strategy, B. Taylor and G. Wills, eds. Princeton, NJ: Brandon System Press, 70-74.

Thaler, R. (1985), "Mental Accounting and Consumer Choice," Marketing Science, 4, 199-214.

Tversky, Amos and Daniel Kahneman (1986), "Rational Choice and the Framing of Decisions," Journal of Business, 59, 4, 2, S251-S278.

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, 15, 95-110.

Westbrook, Robert A. and Michael D. Reilly (1983), "Value-Percept Disparity: An Alternative to the Disconfirmation of Expectations Theory of Consumer Satisfaction," Advances in Consumer Research, 10, 256-261.

Winer, Russell S. (1988), "Behavioral Perspectives on Pricing: Buyer's Subjective Perceptions of Price Revisited," In: Timothy M DeVinney (ed.), Issues in Pricing Theory and Research, 35-57. Lexington, Mass: Lexington Books.

Winer, Russell S. (1986), "A Reference Price Model of Brand Choice for Frequently Purchased Products," Journal of Consumer Research, 13, 250-256.

Zeithaml, Valerie (1988), "Consumer Perceptions of Price, Quality, and Value: A Means-End Model and Synthesis of Evidence," Journal of Marketing, 52, 2-22.

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