본 연구는 상품 컬러가 구매의도에 미치는 영향을 컬러의 대표적 속성이라 할 수 있는 선호도 및 적합도 관점에서 실증 고찰하고 있다. 여기서 컬러 선호도란 말 그대로 특정 컬러를 선호하는 정도라고 할 수 있으며, 적합도는 컬러가 상품과 맞느냐, 맞지 않느냐 하는 개념이라 할 수 있다. 가설 검정을 위해 본 연구는 인쇄 광고를 사용하였으며 피험자로서 대학생 160명이 참가하였다. 연구 결과 선호도가 높은 컬러와 적합도가 높은 컬러는 각각 선호도가 낮은 컬러 및 적합도가 낮은 컬러보다 구매의도를 높이는 것으로 나타났다. 이러한 선호도의 영향은 구매 관여도가 낮은 상품에서 유의하게 나타났으며, 높은 상품에서 유의하게 나타나지 않았다. 반면 적합도의 영향은 구매 관여도가 낮은 상품에서 유의하게 나타나지 않았으며, 높은 상품에서 유의하게 나타났다.
This study investigates effects of product color on purchase intention, specifically focusing on its preference and fit in print advertising. Color preference refers to the extent to which people like a color and color fit is defined as the extent to which a color evokes meanings(i.e. feelings, thoughts, images) that are fit, congruent, or in harmony with those evoked by a product. Based on the past studies, a total of four hypotheses are addressed in the study; High preferred color will show a greater positive effect than low preferred color on the purchase intention for an advertised product(H1), high fit color will show a greater positive effect than low fit color on the purchase intention for an advertised product(H2), the effect predicted by the H1 will be more evident in low purchase- involvement product condition(H3), and the effect predicted by the H2 will be more evident in high purchase-involvement product condition(H4). Especially, ELM(Elaboration Likelihood Model) was used for the formulation of the H3 and the H4. One hundred and sixty undergraduate students at the college of one of authors participated in the study. The sample aged between 20 and 26 with an average of 23.2 and was nearly 60 percent male. Pretests were conducted to select product categories for the manipulation of purchase-involvement and appropriate colors for each preference or fit condition. Each pretest used an independent sample. Based on the pretests, a digital camera and a USB were chosen for high or low purchase-involvement product condition respectively and black(high preferred and high fit condition), clean red-purple(high preferred and low fit condition), yellow green(low preferred and high fit condition), and dull yellow(low preferred and low fit condition) were selected to be used in the main experiment. Before the experiment, participants were told that the purpose of the study was to analyze consumers` responses to product advertising and that they would watch several ads. The ads consisted of one experimental and four nonexperimental ads which were made newly for the study. In order to control for primacy and recency effects, the experimental ad for the high- or low purchase-involvement product condition was placed third in the sequence of five ads. After watching each ad, participants answered the questionnaire items about the ad including their purchase intention for the advertised product. The purchase intention was measured using three items on a seven-point scale(α = .94). The items took the form ranging from ``definitely would not buy`` to ``definitely would buy``, ``improbably buy`` to ``probably buy``, ``definitely would not recommend the advertised product to my friends`` to ``definitely would recommend the advertised product to my friends``. The items were modified slightly from those used in the studies by Goldsmith and Lafferty(2000) and Till and Busler(2000). For further analysis, all multi- items were averaged. In the case of experimental ad, different from nonexperimental ads, manipulation- check measures were added into the questionnaire. Specifically, the purchase-involvement was measured by three seven-point scale items from Park and Young(1986) and Zaichkowsky`s(1985) studies("The purchase of an digital camera(or USB) is important to me," "I watch advertising for a digital camera(or USB) carefully," and "I spend a lot of time in purchasing a digital camera(or USB)") Each item was anchored by 1 = strongly disagree and 7 = strongly agree. The items exhibited high reliability(α = 96). The color preference and fit were measured using one and three items respectively on a seven-point scale. The former was developed by referring to MacInnis and Park(1991), anchored by 1 = not likable and 7 = likable while the latter was from Kellaris, Cox, and Cox(1993) and MacInnis and Park`s(1991) studies(fit, congruent, matched; 1 = strongly disagree, 7 = strongly agree). The reliability among the items for the fit was shown high(α = .90). At the end of class, all participants were given a debriefing questionnaire that asked them in an open-ended format to report what they thought about each advertised product, if there was ad they had watched before the experiment, and what they believed to be the purpose of the study. The entire procedure took about thirty minutes. Because a significant relationship among purchase involvement, color preference, and color fit was revealed in the analyses for the manipulation checks, we used ANCOVA to test the hypotheses instead of ANOVA. Results showed that the high preferred color had a more positive effect on the purchase intention, compared to the low(4.26 vs. 2.59, F(1, 155) = 37.62, p < .001). Likewise, the high fit color showed a greater positive effect on the purchase intention than the low(4.20 vs. 2.65, F(1, 156) = 44.76, p < .001). This implicated that the H1 and the H2 were supported. Also, the effect predicted by the H1 was shown more evident in the low purchase-involvement product condition(high purchase involvement product condition: 4.50(high preference) vs. 3.00(low preference), F(1, 77) = 2.33, N.S.; low purchase involvement product condition: 4.02(high preference) vs. 2.18(low preference), F(1, 77) = 51.07, p < .001) while the effect predicted by the H2 was revealed more evident in the high purchase-involvement product condition(high purchase involvement product condition: 4.78(high fit) vs. 2.71(low fit), F(1, 77) = 144.02, p < .001; low purchase involvement product condition: 3.61(high fit) vs. 2.60(low fit), F(1, 77) = .05, N.S.), which suggested that the H3 and the H4 were supported. The study has several limitations. First, this study used two product categories in examining the effects of product color in advertising. Further research should include other product categories to investigate the extent to which preference and fit of product color substantively influence the product intention. Second, our sample was limited to one population group such as college students. This decreases the generality of the study`s results. Further research should explore the effect by including other population groups and get the confirming results. Third, we performed the study in a restricted situation using nonprofessional ads. It is necessary for further study to examine the effect of music in a real situation using professional ads for more practical implications. Fourth, differently from the expectation, we found a significant relationship among purchase involvement, color preference, and color fit. It would be meaningful if further research find out the reason why such a relationship is formed as well as use more general test methods based on the perfect manipulation-check.