Free Web Hosting Provider - Web Hosting - E-commerce - High Speed Internet - Free Web Page
Search the Web


Framing Effects   Memory and Illusions   Information Processing   Fuzzy-trace theory  Discussion

Go to Previous PageHome PageGo to Next Page

Prospect Theory was put forth by Tversky and Kahneman to understand Framing Effects.  According to prospect theory, people value a certain gain more than a probable gain with an equal or greater expected value; the opposite is true for losses.  Gains and losses are evaluated from a subjective reference point.  The function relating the subjective value and the corresponding losses is steeper than that for gains.  As a result, the displeasure associated with the loss Prospector - Stinkie Peteis greater than the pleasure associated with the same amount of gains.  Therefore, people respond differently, depending on whether the choices are framed in terms of gains or in terms of losses.   The most famous and robust example of framing effects was illustrated by Tversky and Kahnemans’ (1981) Asian disease problem.

In prospect theory, the S shape of the value function predicts framing effects: concave for gains, and convex for losses (see Figure 1, below).  Outcomes are termed in gains and losses, where gains indicates risk aversion and convex for indicating risk seeking.  Framing is a perceptual phenomenon similar to visual illusions, as mentioned in the Memory and Illusions section of this web site.  With framing, outcomes are viewed from two (or more) perspectives, but the objective outcomes remain unchanged.

Top of Page
However, sometimes researchers mistake a subject's choice reversal as a framing effect.  The subject was not necessarily changing their choice due to the framing of a problem, but instead decided the alternate choice was more preferable.  This distinction is not apparent in Prospect Theory.  Fagley (1993) stressed the importance of distinguishing between a choice reversal (reflection effect) and a framing effect when reporting results of framing studies along with identifying the conditions under which either effect can be expected to be observed (Highhouse & Paese, 1996).  Therefore, I believe it is important to make this distinction here.  Framing effects are perceptual.  They are analogous to optical illusions in terms of whether the glass is half full or whether the glass is half-empty.  By phrasing the same outcomes as though they were gains versus phrasing them as though they were losses resulting in different choices made by the subject are considered the framing effect, not the reflection effect. Reflection effects are not as complicated as framing effects.  The reflection effect refers to having a preference for a gamble that is positive, phrased in terms of outcome gains, and then having an opposite preference for a gamble that is negative, phrased in terms of losses (Fagley, 1993).

Figure 1.  Prospect Theory S-Curve
Prospect Theory S-Curve
Source (slight editing - color):

Further distinctions to be made with respect to framing effects were initiated by Wang (1996).  Wang found that framing effects appeared to take two distinct forms.  From this conclusion, he introduced two types of framing effects, 1) bi-directional framing effects, and 2) unidirectional framing effects.

Bi-directional Effects.  This form of the framing effect involves preference reversal from predominantly risk averse to predominantly risk seeking or vice versa, due to the dichotic effect of the framing of the choice outcomes.  This bi-directional framing effect is characterized by predominant risk-averse choices under positive framing and predominant risk-seeking choices under negative framing (Wang, 1996).  This is similar in context, to the standard framing identified through fuzzy-trace theory, which will be explained shortly (Reyna & Ellis, 1994). This could be prevalent in children or adults who are not very much interested, or do not have high levels of personal involvement in a decision.  When these occur, individuals appear to be more vulnerable to framing effects.

Top of Page
Unidirectional Effects.  This form of the framing effect involves no preference reversal but a shift to a more extreme risk preference.  If the predominant preference is uni-directionally risk averse under both framing conditions, it is even more risk averse under positive frame than under negative frame and vice versa.  There are two possible forms of unidirectional framing effect, one augments the risk-averse preference and the other augments the risk-seeking preference. Research in framing effects with high levels of personal involvement has concluded that unidirectional effects occur rather than bi-directional effects (Fagley, 1993; Levin et. al., 1988; Mano, 1994; Paese et. al., 1993; Reyna & Brainerd, 1995; and Wang, 1996).  Studies have shown that with higher levels of personal involvement, bi-directional framing effects are either diminished or even eliminated, but unidirectional effects persist.  This could be that information with high personal involvement has been preserved in exact forms of inputs and therefore, decisions can be made without the evidence of bi-directional framing effects.

Unidirectional effects also appear in decision making problems when groups rather than individuals have been asked to make a decision (Paese, Bieser, & Tubbs, 1993).  When risk seeking individuals were asked to make a decision as a group, their tendency towards risk seeking increased.  However, when risk-averse individuals were asked to make a group decision, their tendencies towards averting risk increased.

Overall, studies on framing effects indicated that they were robust (Kuhberger, 1998).  However, results indicated that the overall framing effect between conditions is only moderate in size, and differences of these effects vary considerably across designs and situational tasks (Kuhberger, 1998; Levin et. al., 1988; Reyna & Brainerd, 1995).   Several investigators have argued that the classical framing effects (bi-directional or standard) occur only when ambiguity about a choice problem is high.  For example, the framing effects occur when people are ambiguous in their experiences of consequences of a decision or if the information provided is not complete (Billings & Scherer, 1988; Frisch, 1993; Levin et. al., 1985; Levin et. al., 1986; Levin et. al., 1988; and Wang, 1996).  Levin et. al. found that subjects were significantly affected by the way in which survey information was framed in rating the incidence of cheating which is consistent with framing effect studies.  However, the effect of the information frame depended on the nature of the tasks.  In their study, Levin et. al. (1988) found that the disappearance of framing effects with increased levels of personal involvement was apparently due to a discounting of the information passage containing the framing manipulation.  They concluded that when a decision is either personal or moral in nature, judges are apparently able to ignore irrelevant outside sources of information and rely on their personal values (Levin et. al., 1988).  In these situations classical framing effects were diminished or eliminated.

Top of Page
On the other hand, Levin et. al. (1986) found that when subjects were not provided with adequate information, increased evidence of framing effects existed.  They found that when subjects were provided with gambles with missing probability information these gambles were selected less often than when paired with gambles where probability was framed in positive terms than when paired with gambles where probability was framed in negative terms.  Subjects were not apt to choose a gamble in which information about the amount to be won was missing, regardless of whether they knew the probability of winning or the probability of losing.  In addition, Reyna and Brainerd (1991) discovered that framing effects were most prevalent when all of the numerical information was replaced by linguistic versions of amount, such as ‘some,’ ‘few,’ ‘many,’ or ‘higher.’  The replacement of the numerical information with these phrases was thought to have made the choices even more ambiguous, hence the increased findings of framing effects.

Wang (1996) also found similar results when he investigated framing effects. Based on his findings he suggested that the framing effects may have resulted from the lack of clarity in choice preferences.  A decision maker with an ambiguous or ambivalent risk preference may actively search for more information besides the tasks, content, and context variables embedded in decision problems.  In this condition, the decision maker’s risk preference may rely on not only the choice options themselves but also the way in which these choice options are worded, phrased, or framed.  Gingrich and Soli (1984) found that the whole process of making a decision in framing problems entailed successively evaluating several alternatives and the results of each evaluation must be remembered and compared.  They believed this procedure placed heavy demands on working memory, and errors, as measured from the linear programming optimal, resulted.  When information presented to the respondent is inadequate, then even higher levels of demands are placed on the working memory and errors are at high risk to occur.

Top of Page
Home Page