713 ADVERTISING STRATEGIES FOR CHARITIES Analysis Study 1 Respondents were recruited from both business and non-business classes at a university in a metropolitan city in Korea. Students were encouraged to participate voluntarily in the experiment. A total of 308 responses were gathered with each cell having 37–40 responses. Care was taken to randomise respondents across cells by mixing up eight types of questionnaire using a random-number table before distributing to respondents. The age range was from 18 to 29 years, with an average of 21.88 years. Males accounted for 46% of the total respondents. An ANCOVA using moral identity and altruism as covariates was run to test the hypotheses. The intention to donate time was subtracted from the intention to donate money, and was used as a dependent variable. Before running a main ANCOVA, the assumption of homogeneity of regression slopes was checked by running an initial ANCOVA that included interaction effects of covariates and independent variables. The initial analysis revealed that all six interaction effects (3 independent × 2 covariates) were non-significant (p > 0.10), confirming the assumption of homogeneity of regression slopes. Table 1 presents the main results. The first hypothesis concerns the effects of appeal type on attracting the two resources. The results showed that the altruistic appeal attracts the intention to volunteer more than it attracts the intention to donate money (M = –0.26 (s.d. = 1.34) ), and the egoistic appeal attracts monetary donations more effectively than it attracts volunteerism (M = 0.45 (s.d. = 1.37) ), and the difference was statistically significant (F = 23.03. p = 0.000). The second hypothesis was also confirmed. Showing needy beneficiaries indicated a larger increase in the intention for monetary donations (M = 0.22 (s.d. = 1.57) ), and presenting helped beneficiaries indicated a larger increase in the volunteering intention (M = –0.02 (s.d. = 1.2) ), and the difference was statistically significant (F = 3.93, p = 0.048). However, showing statistical and identifiable victims made no difference in the
