Aims: We employed a discrete choice experiment (DCE) to investigate students’ preferences for mobile telecommunication plans in a South African University. Study design: Locally optimal DCE were constructed for our choice sets using readily available ideas from blocked fractional factorial designs. This is in contrast to other approaches that may be more complex practically especially when the number of attributes is large. Place and duration of study: The study was conducted in August 2017 at the University of KwaZulu-Natal, Edgewood Campus, Pinetown, South Africa. Methodology: Four hypothetical mobile telecommunication companies were considered and the selected attributes were call rate, data speed, customer service, premiums and network coverage. A two-stage sampling technique was used to select 180 respondents from the student population and data were collected via face-to-face interview. A  blocked fractional factorial design in  blocks each of size four was used to generate the choice sets used to obtain information from the respondents. An extra choice set was included to ascertain the consistency of the choices. Proportion of rational respondents was computed. Multinomial logit model was used to analyze the data and marginal willingness to pay estimates was obtained for the attributes. Results: The proportion of “rational” respondents was 74%. At 0.1% level of significance, the students valued all the attributes except data speed in the process of choosing a particular mobile network. Furthermore, marginal willingness to pay estimates showed that students preferred to pay 51 cents more per minute to have very good customer service. They also preferred to pay extra 13 and 45 cents per minute for more premiums and better network coverage respectively. Conclusion: The results provide empirical evidence of what students perceive as the most important factors influencing their choice of mobile network service providers and these may have decision-making implications for South African-based telecommunication companies.

No abstract