The COVID-19 pandemic has posed a huge challenge to the world in recent years. The development of vaccines that are as effective as possible and accessible to society offers a promising alternative for addressing the problems caused by this situation as soon as possible and to restore the pre-epidemic system. The present study investigated the preferences of residents in Hungary’s second-largest city (Debrecen) for the COVID-19 vaccine. To achieve this aim, a discrete choice experiment was conducted with 1011 participants, and the vaccine characteristics included in the design of the experiment were determined by qualitative methods and a pilot survey: (1) country of origin; (2) efficiency; (3) side effect; and (4) duration of protection. During the data collection at three vaccination sites, respondents were asked to choose between three vaccine alternatives and one “no choice” option in eight decision situations. Discrete choice model estimations were performed using a random parameter logit (RPL) specification with the final model extended to include a latent variable measuring pandemic awareness. The results showed that the vaccine with a Chinese country of origin is the least preferred among the respondents, while the Hungarian and the European vaccines are the most preferred. Furthermore, the increase in the vaccine efficiency level increased the respondents’ sense of utility for the vaccine; the short-term side effect was preferred to the long-term one; and the increase in the duration of protection provided by the vaccine increased the respondents’ sense of utility for the vaccine. Based on the parameter estimated for the latent variable, it can be concluded that as the level of pandemic awareness (which is more positive among people with chronic diseases and less important among health workers) increases, the choice of a vaccine option becomes more preferred among respondents compared to the “no choice“. The results of our investigation could contribute towards increasing compliance in the case of the vaccination-rejecting population, not only for COVID-19, but for any kind of vaccination procedure.

Heterogeneity in preferences can be addressed through various discrete choice modeling approaches. The random-parameter latent class (RLC) approach offers a desirable alternative for analysts due to its advantageous properties of separating classes with different preferences and capturing the remaining heterogeneity within classes by including random parameters. For latent class specifications, however, more empirical evidence on the optimal number of classes to consider is needed in order to develop a more objective set of criteria. To investigate this question, we tested cases with different class numbers (for both fixed- and random-parameter latent class modeling) by analyzing data from a discrete choice experiment conducted in 2021 (examined preferences regarding COVID-19 vaccines). We compared models using commonly used indicators such as the Bayesian information criterion, and we took into account, among others, a seemingly simple but often overlooked indicator such as the ratio of significant parameter estimates. Based on our results, it is not sufficient to decide on the optimal number of classes in the latent class modeling based on only information criteria. We considered aspects such as the ratio of significant parameter estimates (it may be interesting to examine this both between and within specifications to find out which model type and class number has the most balanced ratio); the validity of the coefficients obtained (focusing on whether the conclusions are consistent with our theoretical model); whether including random parameters is justified (finding a balance between the complexity of the model and its information content, i.e., to examine when (and to what extent) the introduction of within-class heterogeneity is relevant); and the distributions of MRS calculations (since they often function as a direct measure of preferences, it is necessary to test how consistent the distributions of specifications with different class numbers are (if they are highly, i.e., relatively stable in explaining consumer preferences, it is probably worth putting more emphasis on the aspects mentioned above when choosing a model)). The results of this research raise further questions that should be addressed by further model testing in the future.