Research Design
The factorial design is used to measure two or more variables and to evaluate the main effect of each individual
variable and the interaction effect between them. As for experimental units, since students share much more in
common, and females are usually the primary users of cosmetic products, this research targets female students at
a university. Wearing make-up has become a social courtesy. According to statistics, women with the ability to
consumption spend about NT$10,000 purchasing cosmetic products each year, and beauty products take up
almost half of the sales volume. The annual cosmetic sales also increase overall sales volume of retailers. Thus,
cosmetic beauty products have been chosen as the experimental product in the research to see if brand image and
product knowledge affect purchase intention and if price discount creates any moderating effect. This research
includes two pre-tests and one formal test. Pre-test 1: Cosmetic beauty products can be divided into three main
parts: beauty products, cosmetics and perfume. Pre-test 1 aimed to determine which type of product is the most
commonly purchased and the overall impression consumers have towards cosmetic beauty product brand image.
Thirty questionnaire were sent out to female students majored in business administration at the university. The
results showed that Christian Dior received the highest average score of 5.60, while Avon received the lowest
score. The average score for the others was 4.30. Since there are significant paired test differences(t=5.281,
p=0.000)between the highest and lowest brand images, the two cosmetic beauty brand names, Christian Dior
and Avon, were selected as two experiential targets.