7. Discussion Our findings on the differential impacts of CoI dimensions across product classes suggest that CoI effects should be investigated at the dimensions’ level (not as aggregated effects of a more abstract construct) and that they may not be fully compensatory (as also argued by Wang, Li, Barnes, & Ahn, 2012). Similarly to Witt and Rao’s (1992) findings, our results show that the overall CoI effect (as indicated by R2 in the columns about the differences approach in Table 3) is smaller on clothes than on home appliances (or on fruits in general, for that matter), but the differences were not statistically significant, at least in our sample. Additionally, similarly to Eroglu and Machleit (1989), who found that the CoO effect was higher on technologically advanced products, our results indicate that the overall explained variance of product evaluation of home appliances (a more advanced product; explained variance equal to 41%) is higher than that of fruits (25%) or that of clothes (16%). However, our findings challenge Brijs et al.’s (2011) contention that the effects of CoI would be stronger for hedonic-oriented than for utilitarian-oriented products.
