90 / 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
90
condition but less strongly than in the SE condition. In brief, all of the findings prove that
people tend to use a heuristic process in SE modes, but a systematic process in SQE and JE
modes.
Finally, a possible explanation for the insignificance of the perceived quantitative
estimation of the battery stimulus is that the maximum number that could be purchased
within a given budget was relatively smaller (14.4 items for a $2.99 battery) than that of the
other two stimuli (23.6 items for $1.99 nail clippers and 32.6 items for a $6.99 baseball cap).
This made its quantitative estimation comparatively a more cognitive process than that of the
others, which diminished the nine-ending pricing effect.
3.3 Study 3
Study 2 demonstrated that influence of JE and SQE modes were similar on the nine-
ending effect. Toward this end, the aim of Study 3 was designed to examine the nine-ending
pricing effect from various perspectives between SE and JE modes. According to the left-
digit effect, when people encounter multi-digit prices they pay most attention to the leftmost
digit, which has a more dominant influence on their magnitude perception than the other
digits to its right. It follows that the nine-ending pricing effect is more likely to be
manifested when the nine-ending price and zero-ending price one dollar higher have
different leftmost digits (e.g., $199 vs. $200). While previous studies have adopted different
leftmost digits between nine-ending price and zero-ending prices, this research sets these
leftmost digits at the same value (e.g., $129 vs. $130). This diminishes the dominance of the
leftmost digit in price perception when comparing nine-ending and zero-ending prices one
dollar higher.
Since the difference in magnitude perception is smaller between nine-ending and zero-
ending prices with the same leftmost digits than between those with different leftmost digits,
without the anchoring of the leftmost digit, the nine-ending and zero-ending prices are
allocated onto locations on the mental analog magnitude scale (Dehaene, 1997; Dehaene,
Dupoux, and Mehler, 1990) not far from each other. As a result, the influence of the nine-
ending pricing effect diminishes since nine-ending prices are no longer perceived to be much
smaller than zero-ending prices one dollar higher, regardless of the way they have been
evaluated (separately or jointly). When the leftmost digits are the same for both nine-ending
and zero-ending prices, the nine-ending pricing effect diminishes to the extent that the
difference in the perceived price magnitude between these two kinds of prices is viewed as
insignificant in both separate and joint evaluation conditions.