ABSTRACT
In this study, we leverage valence theory, cognitive absorption theory, and IT adoption literature to investigate the perceptions of consumers towards the use of online Recommendation Agents (RAs) that vary in the number of details they provide in eliciting consumers’ preferences and presenting recommendations accordingly. The research model is empirically validated via an experiment involving 197 online shoppers. Results show that high in-depth RAs are better alternatives to low in-depth RAs in driving consumers’ intention to use RAs in their shopping experience. The findings provide novel insights for researchers and practitioners interested in understanding the proper design for online RAs.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1 The standard deviations and means of constructs are also shown in Appendix A (see ).
2 Participants’ interest in purchasing a car was considered as 1: very low, 2: low 3: some, 4: high, 5: very high; user RA experience was considered as 1: very low, 2: low 3: some, 4: high, 5: very high; education level was considered as 1: high school, 2: college diploma, 3: bachelor's degree, 4: master's degree, and 5: PhD degree; Web experience was considered as 1: less than 1 h, 2: between 1 and 5 h, and 3: more than 5 h; and task complexity was considered as 1: very low, 2: low 3: some, 4: high, 5: very high.
3 Results also showed that the control variables (i.e., task complexity, internet use, education level, RA experience, car purchasing interest) had no impacts on dependent variables.