94 / 372
九尾數定價效果在不同評估模式中的侷限：分別、聯合與依序評估
94
proposed (Bettman, Johnson, and Payne, 1990; Bettman and Kakkar, 1977; Campbell, 1988;
Jiang and Benbasat, 2007). However, decision effectiveness begins to decline until task
complexity reaches a reversed point induced by an individual’s limited cognitive capacity
(Swait and Adamowicz, 2001). In sum, cognitive loading is an important factor in decision
switching (Garbarino and Edell, 1997).
In view of these studies and the HSM, this paper maintains that if the increasing
demand on cognitive capacity induced by greater task complexity exceeds what people may
possess, there will be an increased tendency to use a heuristic process that will lower
cognitive effort and response time. In other words, the difference in the perceived price
magnitude between nine- and zero-ending digits will keep on increasing in the JE mode
when the task becomes more and more complex. No nine-ending effect will occur both in the
SE and JE modes when the degree of task complexity exceeds an individual’s cognitive
loading because people will just make a rash judgment. Following the study of Thomas and
Morwitz (2005), Study 4 tries to explore the bounded effect of nine-ending prices with
increasing degrees of task complexity in the different evaluation modes and measure the
response time so as to confirm which processes people tend to use. The following hypothesis
is proposed:
H5: The greater the task complexity, the less cognitive effort is expended in the JE
mode, and the less difference in the perceived price magnitude between the SE and
JE modes.
3.4.1 Method
One hundred undergraduate students (sixty-one males and thirty-nine females) were
recruited from a large southern university. The stimuli were the same as in Studies 1, 2, and
3. There were two items of perceived price magnitude within each product category, each
with different fictional brand names (nail clippers: S and K; battery: R and T; baseball cap: P
and U) and ending prices. The prices of the stimuli were manipulated to have a nine-ending
price with a lower value in the leftmost digits (S: $1.99; R: $2.99; P: $6.99), whereas the
prices of the K, T and U-branded nail clippers were manipulated to have a zero-ending prices
with higher values in the leftmost digits ($2.00, $3.00, and $7.00). This study utilized a
between-subjects design with one factor (price magnitude) and five-levels (separate: S, K;
joint: S-K, S-K-R-T, S-K-R-T-P-U). The participants were randomly divided into eight
subgroups and were only asked to state the perceived price magnitude of S and K. In order to
resolve the order effect, a counterbalance design was used. For example, S-K-R-T and R-T-