Choosing by Liking (Affect Heuristic)

Choosing by Liking

Choice is typically modeled as a cognitive procedure involving an analysis of an option’s constituent features. However, some have proposed an alternate view, arguing that choices might also be made intuitively; by the spontaneous affective evaluation of liking or disliking that options may elicit – a procedure that Wright (1975) calls affect referral, Schwarz and Clore (1988) term the “How-do-I-feel-about-it” heuristic, and Slovic, Finucane, Peters, and MacGregor (this volume) call the affect heuristic. Image theory (Beach & Mitchell, 1987; Beach, 1990) advanced a similar view, proposing that affective evaluation may work in conjunction with cognitive evaluation – that affective valuation is used as a quick initial screen for alternatives, whereas cognitive evaluation is reserved for those alternatives that surpass some affective threshold.

Relying on intuitive affective impressions (what Kahneman terms choosing by liking) may be a successful heuristic if the features that mediate our initial affective response correspond closely with the features that determine our subsequent enjoyment. If the spaciousness of apartments determines how much we like them while apartment hunting and determines how much we actually enjoy living in them, choosing by liking may be a perfectly adequate heuristic. However, if affective evaluation is unconsciously governed by something not closely correlated with subsequent satisfaction (e.g. similarity to our previous apartment, the song that is playing when we enter), choosing by liking could lead one astray.

In general, the reliance on spontaneous affective response may be an effective decision heuristic. Indeed, Wilson and colleagues have suggested that affective impressions may be more accurate than analytic assessments of an option’s constituent features – that a more thorough cognitive analysis can degrade rather than enhance evaluative validity. To this effect, Wilson et al. (1989) quote a judge at the Berlin Film Festival who noted:

I went to every screening with a fresh pack of note cards that I would dutifully cover with my impressions of each and every film. The result, of course, was that the movies ceased to be fun and turned into problems, a struggle against time, darkness and my own esthetic emotions which these autopsies confused. I was so worried about evaluating every aspect of every film that my entire system of values went into shock, and I quickly realized that I could no longer easily tell what I liked or didn’t or why (p. 288; originally published in Vargas Llosa, 1986, p. 23).

Some experimental results seem to confirm this observation. For example, Wilson and Schooler (1991) found that subjects who simply rated how much they liked various brands of strawberry jam produced ratings that corresponded better with gustatory experts than subjects who first listed the reasons they liked or disliked a jam before rendering their global ratings. Similarly, Wilson et al. (1993) found that subjects who were instructed simply to choose the poster they liked most ended up liking their poster more (and were more likely to still be sporting it on their dormitory wall) than subjects who first analyzed each poster along several dimensions before making their final decision. When people reflect about the reasons they like or dislike something they may focus on features that seem to be plausible determinants of liking rather than the features that actually mediate liking.

Although affective impressions may be often an accurate proxy for the overall quality of an option, and superior to clumsy attempts at multiattribute utility analysis, several properties of affect render it an imperfect and, sometimes, markedly deficient basis for choice: (1) it is insufficiently sensitive to quantitative detail; (2) it is unduly influenced by transient contextual cues; (3) it is excessively affected by familiarity.