臺大管理論叢 NTU Management Review VOL.26 NO.2

臺大管理論叢

第

卷第

期

conditions of nine-ending effects, and to extend the research to incorporate different

evaluation modes. Experimental results are presented and concluding remarks are provided

together with implications and directions for future research.

2. Literature Review

2.1 Evaluation Modes: Separate, Joint and Sequential

The Evaluability Hypothesis maintains that it is more difficult to evaluate the

desirability of values associated with certain attributes than those associated with others, and

that compared to easy-to-evaluate attributes, difficult-to-evaluate attributes will have a

greater influence on joint evaluation (JE) than on separate evaluation (SE) (Hsee, 1996;

Hsee, Loewenstein, Blount, and Bazerman, 1999). As these terms imply, SE refers to the

condition where two stimulus options are separately presented and evaluated, whereas JE

refers to the condition where two stimulus options are presented and evaluated side by side at

the same time (Goldstein and Einhorn, 1987). These evaluation modes are utilized to argue

against normative decision theories, which suggest that regardless of the way preferences are

elicited, people’s preferences remain unchanged; in reality, people may display different or

even reverse preferences within two normatively equivalent evaluation conditions (Hsee,

1996; Schmeltzer, Caverni, and Warglien, 2004). Hence, the preference for stimulus options

changes according to the evaluation condition.

For example, Hsee (1996) has designed a second-hand dictionary scenario to examine

the amount that people say they are willing to pay. Dictionary A contains 10,000 entries and

looks new; Dictionary B contains 20,000 entries and has a torn cover. The results show that

people will assign a higher price to dictionary A ($24) than to dictionary B ($20) when they

make their judgment in the SE mode. However, in the JE condition, the opposite results

occur: people assign a higher price to dictionary B ($27) than to dictionary A ($19). In this

example, the number of entries is regarded as a difficult-to-evaluate attribute in the SE mode;

thus, the easy-to-evaluate attribute, “new” versus “old,” has a greater influence on the

people, who are more willing to pay for Dictionary A than Dictionary B. Conversely, as the

Evaluability Hypothesis

proposes, people will compare the difficult-to-evaluate attribute in

2 Although researchers use Evaluability Hypothesis to explain the comparisons among two or more attributes,

it is reasonable to compare the single attribute of price in this article. Based on prior studies, price may play

a dual role in consumers’ product evaluations at the same time. For example, Bornemann and Homburg

(2011) present that consumers may interpret price information as either an indicator of quality or an

indicator of monetary sacrifice when evaluating a product. That is, the extension of Evaluability Hypothesis

is suitable and acceptable here.