Perspectives on Framing

Gideon Keren

• Publisher: Psychology Press
• Year: 2011
• eBook ISBN: 9781136988646
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

Language comprises a major mark of humans compared with other primates and is the main vehicle for social interaction. A major characteristic of any natural language is that the same communication, idea, or intention can be articulated in different ways—in other words, the same message can be "framed" differently. The same medical treatment can be portrayed in terms chance of chance of success or chance of failure; energy reduction can be expressed in terms of savings per day or savings per year; and a task can be described as 80% completed or 20% uncompleted. In this book, contributors from a variety of disciplines—psychology, linguistics, marketing, political science, and medical decision making—come together to better understand the mechanisms underlying framing effects and assess their impact on the communication process.

Part I
Theoretical Perspectives

1
On the Definition and Possible Underpinnings of Framing Effects: A Brief Review and a Critical Evaluation

GIDEON KEREN

Language is one of the more important aspects that distinguish human beings from other primates. While it plays a fundamental role in cognition, affect, and social psychology, the exact nature of its relation with these domains is not yet entirely understood (Fiedler, 2008). Specifically, the complex link between language and cognition (and the interplay between them) is an old question that has never been entirely elucidated. Is language a reflection of the internal representation of our cognitive system, or does it also actively affect it? Research on framing – the central theme of the present book – suggests that the relation may work in both directions.

The term frame is associated with multiple meanings. In its most abstract interpretation it implies the composition of different parts of a message according to a particular design. It can refer to a construction (e.g., frame of a building), to a surrounding or a border (e.g., frame of a picture), to a state of mind (e.g., she is in a happy frame of mind), or to the linguistic composition of a sentence or an utterance. What all these (and related) usages have in common is that they afford a certain structural basis or, in perceptual terminology, determine the Gestalt of the message. The current book centers on linguistic and psychological aspects of framing, examining the consequences of employing different message frames.

Message framing is a topic that has attracted researchers from different disciplines, including linguistics, psychology (in particular, behavioral decision making), communication, economics, marketing, politics, and health psychology, and is in fact relevant for any domain in the social sciences. The subject is important for both theoretical and applied reasons. From a theoretical perspective, it provides a challenge to the concept of rationality as well as to economic theory (e.g., Kahneman, 2000; Kahneman & Tversky, 1984; Tversky & Kahneman 1981, 1986; but see Sher & McKenzie, 2006, 2008; and Chapter 2). In daily life it may have far-reaching consequences on decisions such as choice of a medical treatment (see Chapter 12), selection by expert financial planners of one out of several alternative investments (Roszkowski & Snelbecker, 1990), and the extent to which one may be willing to volunteer as an organ donor (Johnson, Bellman, & Loshe, 2002). It may also have important implications for assessing and interpreting people’s opinions in different sorts of surveys and opinion polls (see Chapter 14), depending on how different questions are formulated and presented.

Framing, even if not framed in this term, has a rich history dating back to the study of rhetoric (e.g., Giambattista Vico, 1668–1744). More recently, the sociologist Goffman (1974) presented his book entitled Frame analysis, which contains a broad theoretical approach used in communication theory. Introducing his book, he suggested that it “is about the organization of experience – Something that an individual actor can take into his mind” (p. 9). The work of Kahneman and Tversky (1984; Tversky & Kahneman, 1986) served as a most important impetus to a recent stream of systematic research on framing effects. Indeed, much of the research described in this volume can be traced back to their demonstrations of framing effects, the most notable one being the so-called “Asian disease” problem.

Despite the intensive research on the topic, the processes underlying the growing number of demonstrations of framing effects are far from clear. Recently, Sugden (2005) proposed the distinction between experiments as exhibits (with the purpose of demonstrating the existence of a phenomenon) and experiments as explanations (with the purpose of testing a given explanation). Most framing studies seem to fall under the former rather than the latter category.

The purpose of this chapter is to briefly examine the knowledge that has been accumulated on framing effects, and at the same time raise some queries that still have to be addressed. I start by discussing several definitions proposed in the literature and critically examine some of the common typologies of framing effects. I then explore some theoretical frameworks that attempt to provide a broad account of different framing effects, and highlight the different aims and methodologies adopted by different approaches. I close the chapter by pondering whether a unified theoretical framework of framing is possible, or whether the subject will remain a fragmented research program.

ON THE DEFINITION OF FRAMING

Generally, one may distinguish between a loose and a strict definition of framing. According to the loose (broad) definition, framing concerns different linguistic descriptions of the same problem – different formulations of the same message – that transmit the same core information. I label this characterization as the loose view because what constitutes the “core” information remains ambiguous. Following the loose perspective, framing concerns the way the story is written or produced, and correspondingly the meaning that evolves. Different frames shape the meaning underlying the (presumably same) message, and result in different construals of the internal presentation. The visual counterpart of framing is a painting placed in different frames, yielding somewhat different perceptions of the same painting. Framing, following the loose perspective, can be achieved through the use of different articulations attained by highlighting different aspects, referred to by Druckman (Chapter 13) as emphasis framing, as long as the main message does not change. Accentuating different elements of a message may convey some additional (specific) information and consequently yield different interpretations.¹ The loose approach has no well-defined criteria as to when two frames should be considered comparable. Any two frames are acceptable as long as they convey sufficiently similar information (similarity in this context is not well defined and is hard to measure).

Following a more restrictive characterization, two frames constitute a re-description of exactly the same message or situation if they are logically equivalent. For instance, in the well-known Asian disease problem (Tversky & Kahneman, 1981),² the consequences of a rescue program formulated as “200 out of 600 will be saved”, or “400 out of 600 will die”, are assumed to be equivalent. A visual analog to logical equivalence can be found in the so-called reversible or ambiguous figures, such as the Necker cube, in which the same stimulus can be interpreted in more than one way. The link between verbal framing and reversible figures can also be stretched to the conjectured underlying processes. One common explanation for perceiving more than one pattern in the same (reversible) figure is based on the fact that various organizations of the same retinal image are possible (Rock, 1975), which experientially correspond to different figure–ground relationships. Similarly, numerous verbal framing effects can be explained in terms of a shift in the focus of attention. Following this explanation, different frames are interpreted differently because they represent different perspectives of the same scene – the situation is viewed from different angles, and compared with different reference points. The reference point in this context is the analogue of the background in the interpretation of reversible figures.³

Note, however, that reversible figures represent the ultimate equivalence in that the stimulus (e.g., the Necker cube) that triggers more than one interpretation remains exactly the same. In linguistic framing, even if the requirement of logical equivalence is satisfied, the stimuli that participants are exposed to, namely two different frames, are not identical. Moreover, fulfillment of the formal (normative) requirement of logical equivalence does not necessarily ensure that it will also be perceived as such. For instance, when 600 people are expected to die, the statement “200 will be saved” seems to be formally equivalent to “400 will die”. Yet, nothing is said about the remaining 400 in the former frame or the remaining 200 in the latter frame (Mandel, 2001). Frisch (1993) has shown that a substantial proportion of participants who exhibited framing effects did not agree that the above two frames should be treated as the same. Further, Mandel (2001) claimed that the “sure” options in the Asian disease problem confound two manipulations: one that refers to the expected outcomes that are made explicit (i.e., positive or negative) and one that refers to the descriptors used to convey the relevant expected outcomes (i.e., lives saved/not saved vs. lives lost/not lost). Sher and McKenzie (2006, 2008; and Chapter 2) proposed an even stricter criterion for considering two frames to be alike, namely that they should be informationally, (rather than logically) equivalent. Frames that are logically equivalent might not be informationally equivalent if the speaker chose one rather than another frame with the aim to convey specific informative to the listener.

The above brief discussion suggests that the difference between a strict and a loose approach largely depends on the necessary conditions required for two frames to be considered comparable or equivalent, an issue further discussed below. Naturally, there is a wide range of possibilities between a most restrictive and a more lenient definition of framing. Many researchers avoid discussing in depth the appropriate definition of framing and the corresponding assumptions they embrace. Which definition is adopted is essential for interpreting the empirical results. In addition, the chosen definition would determine to which other framing research the results could be linked. Most important, there is an unavoidable tradeoff between a narrower (stricter) definition that is amenable to the development of a coherent theoretical framework, and a broader one that may offer more generality and contain a broader set of (so-called) framing effects but at the price of ending up with a fragmented research domain with little, if any, theory.

One other interpretation of framing, which has received relatively little attention, is based on what Druckman refers to as frames in mind (see Chapter 13). Students of perception are familiar with the issue of figure–ground perception, but the perception and interpretation of an object or event are determined not just by the external but also by the internal background – by the internal cognitive structure and current existing knowledge. Thus, a message may sometimes be interpreted in the same way, regardless of how it is framed, due to a rigid existing mental structure (or mental model) that is narrowly tuned to a specific world view. The history of science is filled with examples where factual evidence was distorted in order to be adjusted to an accepted view (Polanyi, 1958), the most well known being the time it took to accept the Copernican “mental model” in favor of the Ptolemaic (which persisted for centuries). A full understanding of framing would require an examination of the interaction between what Druckman termed frames in communication (the linguistic aspects of a frame, such as wording, structure, and presentation style) and the frame of mind on which the message is impinged.

Kahneman (2000) noted that the choice of the term frame to descriptions of decision problems may be unwanted because it can be interpreted at two levels: at one level a frame is the linguistic formulation of a message to which listeners (decision makers) are exposed; at another level it refers to the listeners’ construal of the message. Thus, following Kahneman, framing is used to describe both an experimental manipulation as well as a constituent activity of the listener (the decision maker). Note that frame of mind refers to how the initial mental model (before receiving the message) effects the interpretation of a frame. In contrast, when framing is viewed as the listener’s construal of the message, it is the frame that influences the mental model.

Clearly, the different facets of framing implied from the above discussion are deeply interrelated, suggesting that the concept is complex and does not have a single unique definition. To simplify things, and supposedly in order to offer some further insights into framing effects, several framing typologies have been proposed that are briefly discussed in the following section.

ALTERNATIVE CLASSIFICATIONS OF FRAMING EFFECTS

To surmount the thorny problems associated with defining framing, several typologies have been introduced in the literature. Often, when reporting demonstration of framing effects, the findings are delineated by placing them explicitly in a specific category within one or the other typology. The value of any typology of framing effects would seem to depend on three issues:...

Table of contents

• Contents
• About the Editor
• Contributors
• Preface
• Part I Theoretical Perspectives
• Part II Linguistic Aspects of Framing
• Part III Framing Effects: Empirical Findings
• Part IV Applications
• Index