There is a significant difference in the number of references to physical characteristics by country (p = 0.029) and sex (p < 0.001); there is no two-way interaction (p = 0.163). Gender differences are as predicted (H1b); however, H2b receives only partial support. Australians were more prone to mention physical characteristics than were Indi­ans, but those from Hong Kong had the greatest proclivity overall.

With respect to entertainment services men­tioned, there is a significant difference due to country (p < 0.001); and the predicted gender effect only approaches significance (p = 0.058).

There is, however, a significant gender by country interaction (p = 0.029). Overall, and consistent with H2c, Australians are more likely to mention entertainment services than are individuals from India or Hong Kong. However, an interesting pattern emerges with respect to gender: India and Hong Kong males make more entertainment – related comments than do women, whereas the reverse is true in Australia. H1c is therefore not supported.

Concerning comments about personality traits, there was a main effect due to country (p < 0.001), but no significant difference due to gender (p = 0.392) or the two-way interaction (p = 0.367). These data support H2d: Australians are more prone to mention personality traits than are
individuals from India or Hong Kong (and, as it turns out, Indians are more prone than are those from Hong Kong).

There is a significant difference in the num­ber of references made about educational status by country (p < 0.001); however, there is no significant gender effect (p = 0.358) nor two­way interaction (p = 0.778); H1e is therefore not supported. Indians made more education-related references than did individuals from Hong Kong or Australia, although there was no significant difference between the latter two; H2f therefore receives only partial support.

With respect to comments about one’s intellect, there was as a significant effect due to country (p = 0.011), but not a significant effect due to gender (p = 0.384) or the two-way interaction (p= 0.384), H1f is therefore not supported. Furthermore, con­trary to expectations the main effect of country was opposite to the expected pattern: Australians were more likely to make intellect-related references than were individuals from either India or Hong Kong; H2g is therefore not supported.

There is a significant difference in references to occupational status by country (p = 0.004); but no significant difference due to sex (p = 0.662) or the sex by country interaction (p = 0.054); H1g is therefore not supported. Contrary to expectations, Australians were more prone to make occupation – related references than were individuals from India or Hong Kong; H2h is therefore not supported.

With respect to references to demographic information, there is a significant main effect due to country (p = 0.006) as well as gender (p = 0.041); there is also a significant two-way in­teraction (p = 0.036), the latter driven by India. While individuals from Hong Kong were more likely to mention demographic information than were individuals from Australia, the same cannot be said about individuals from India; thus, H2j receives partial support.