For each of the ten categories of interest a 3 x 2 ANOVA was run, the dependent variables being country and gender. In all but two cases (money and ethnic information), the overall model’s F – statistic was significant at p =.05, hence only the significance of the main and two-way interac­tion effects are reported. In the case of money and ethnic information, the number of subjects reporting zero counts (i. e., made no reference to these issues) were n=232 and n=210, respectively, hence these categories were excluded from further analysis. Therefore, H1d, H2e and H2i are reported as having insufficient data, an interesting insight in itself. (Box 1 and Box 2)

We fully elucidate the analysis pertaining to love-related references, the first category of inter- est. Analyses for the other categories were similar, but the findings are presented more succinctly.

Box 1. Australian male coding counts

Love: 3 (appropriate words underlined for illustrative purposes)

Personality trait: 6 (in italics)

Entertainment: 4 (including its obverse, ‘Not much into clubbing’)

Physical characteristic: 1 (in bold)

Box 2. Indian female coding counts

Love: 4 (underlined)

Personality trait: 6 (in italics)

Entertainment: 0 Physical characteristic: 0

Based on the 3 x 2 ANOVA with love-related references the dependent variable, there was a main effect due to country (p <.001) as well as a significant two-way interaction between country and gender (p<.001); however, there was no main effect due to gender (p=.865). Cell means appear in the figure below. Australians were signifi­cantly more likely to make love-related refer­ences (x = 1.92) than were either individuals from Hong Kong (x = 1.09) or India (x =.71). Pairwise comparisons revealed that the main effect of country was due to a significant difference between India and Australia (p <.001), as well as a sig­nificant difference between Hong Kong and Australia (p =.009); however, there was no sig­nificant difference between India and Hong Kong
(p=.123). Findings related to country are therefore consistent with H2a. (Figures 1, 2, 3, 4, 5, 6, 7, & 8)


To examine the two-way interaction, t-tests by gender within each country were conducted. Within India there was no significant difference in the number of love-related references between men and women (p =.133), whereas there was a significant difference between genders in both Hong Kong and Australia (p <.001). Surpris­ingly, men were more likely to make love-related references than were women in Hong Kong (means equal 1.73 and 0.45, respectively), whereas the reverse was true in Australia (mean equals 1.12 versus 2.79). We therefore cannot support H1a.