評估「善」與「惡」組合下的平均偏誤效果與語意定錨效果：體重在意程度之調節效果

Abstract

本研究旨在探討：當體重在意程度較高或較低的人們面臨同時含有兩種與內心目標互相衝突之極具關聯性的食物組合－「善」與「惡」－時，人們是同時看見「善」與「惡」、先看見「善」再看見「惡」或是先看見「惡」再看見「善」的食物組合，他們估計該食物組合之總熱量是否會有顯著的差異。 本實驗之結果顯示，平均而言，人們合理地認為同時看見之「善」與「惡」的食物總熱量高於單一個「惡」的熱量，此結果與本研究假設及過去文獻推論結果相反。其中，體重在意程度較高的人們認為同時看見之「善」與「惡」整體健康印象，比起單一個「惡」還要健康，因而在估計食物的熱量時，認為「善」與「惡」的總熱量低於單一個「惡」的熱量，此結果與本研究假設及過去文獻推論結果一致，儘管明顯違反了理性消費者應有的一致性以及數學加法的邏輯；反之，體重在意程度較低的人們則是很合理地認為「善」與「惡」的總熱量高於單一個「惡」的熱量，此結果與本研究假設及過去文獻推論結果相反。以上結果顯示只有體重在意程度較高的人們才會發生平均偏誤現象。 一旦人們看見「善」與「惡」的食物組合有先後順序的分別，不論是體重在意程度較高或較低的人們皆會認為，有先後順序之「善」與「惡」的食物總熱量高於沒有先後順序（同時看見）之「善」與「惡」的食物組合。平均而言，人們認為先看見「善」再看見「惡」的食物總熱量，低於先看見「惡」再看見「善」的食物組合，此結果顯示過去文獻所證實之語意定錨效果所引發的對比效果並未發生。反之，研究者所觀察到的卻是傳統定錨效果所引發的同化效果，和其它大部分的文獻結果推論一致。其中，就先看見「善」再看見「惡」的食物組合而言，體重在意程度較高的人們所估計的總熱量低於體重在意程度較低的人們所估計的總熱量，相反地，若是先看見「惡」再看見「善」的食物組合，體重在意程度較高的人們所估計的總熱量卻又高於體重在意程度較低的人們所估計的總熱量。 綜合本研究所獲之結論，尤其是對於那些在意自身體重、體態或是正在減重的人們而言，只要一餐中同時擁有「善」與「惡」的食物時，若在評估熱量時逐一分開來評估，就不會令他們過於低估該食物組合的熱量，也就大大降低了可能攝取過多熱量的情形。而若是更進一步將焦點先放在熱量高、「惡」的食物上，爾後再來評估其它相對而言熱量低、「善」的食物，對於正在減重的人們來說或許效果會更好。

This research is trying to figure out when weight-conscious vs. weight-unconscious individuals face food combinations of vices and virtues (which is highly related to indulgent goals and healthy goals conflicting in their mind), are there any significant difference between their evaluation of calories of the combined meal when it is in joint evaluation of vice and virtue, in vice-virtue sequence, or in virtue-vice sequence. The experimental results show that people in average tend to believe the combined meal in joint evaluation of vice and virtue has more calories than the vice alone, and in this case averaging bias effect does not happen. But as for weight-conscious individuals, their evaluation does show the averaging bias effect. They estimate that the combined meal in joint evaluation of vice and virtue has fewer calories than the vice alone. Moreover, as long as the combined meal are sequentially evaluated, both weight-conscious and weight-unconscious individuals consider the calorie content of the combined meal in sequential evaluation are greater than that of the meal in joint evaluation, which corrects the underestimation effect in joint condition. Besides, both kinds of individuals believe vice-virtue sequence of the combined meal has more calories than virtue-vice sequence of it, because these sequentially evaluated alternatives do not lead to contrast effect due to semantic anchoring, but lead to assimilation effect due to numeric anchoring. My conclusion is that especially for those who really care about their weight or body shape, when they face a meal combining vices and virtues, as long as they remember to evaluate the calorie content of these types of food separately and sequentially rather than considering them together, they will not underestimate the combined meal, and will not be unconsciously eating too much. Specifically speaking, if dieters put the main focus on evaluating the calories of the vice first, and then evaluate that of the virtue, their dieting results may well be more effective.