A Two-Stage Model for Factors Influencing Citation Counts

This work aims to use a suitable regression model to study a count response random variable, namely, the number of citations of a research paper, that is affected by some explanatory variables. The count variable exhibits substantial variation, as the sample variance is larger than the sample mean; thus, the classical Poisson regression model seems not to be appropriate. We concentrate our attention on the negative binomial regression model, which allows the variance of each measurement to be a function of its predicted value. Nevertheless, the process of citations of papers may be divided into two parts. In the first stage, the paper has no citations, while the second part provides the intensity of the citations. A hurdle model for separating documents with citations and those without citations is considered. The dataset for empirical application consisted of 43,190 research papers in the Economics and Business field from 2014–2021, which were obtained from The Lens database. Citation counts and social attention scores for each article were gathered from the Altmetric database. The main findings indicate that both collaboration and funding have positive impacts on citation counts and reduce the likelihood of receiving zero citations. Open access (OA) via repositories (green OA) correlates with higher citation counts and a lower probability of zero citations. In contrast, OA via the publisher’s website without an explicit open license (bronze OA) is associated with higher citation counts but also with a higher probability of zero citations. In addition, open access in subscription-based journals (hybrid OA) increases citation counts, although the effect is modest. There are clear disciplinary differences, with the prestige of the journal playing a significant role in citation counts. Articles with lower expert ratings tend to be cited less frequently and are more likely to be cited zero times. Meanwhile, news and blog mentions boost citations and reduce the likelihood of receiving no citations, while policy mentions also enhance citation counts and significantly lower the risk of being cited zero times. In contrast, patent mentions have a negative impact on citations. The influence of social media varies: X/Twitter and Wikipedia mentions increase citations and reduce the likelihood of being uncited, whereas Facebook and video mentions negatively impact citation counts.