*2.3. Data Analysis*

Data extracted by the independent researchers were transferred into STATA (version 14.0, STATA Corp., TX, the USA). Descriptive statistics were used to generate results for the following publication characteristics: (1) the number of publications produced per year, (2) the number of publications per country (from 1991–2018), (3) the average number of citations for all published papers per year, (4) the frequency of publication downloads in the last six months of 2018 and in the period of 2014–2018, and (5) the average number of publication downloads in the last six months of 2018 and during the period of 2014–2018. Latent Dirichlet Allocation (LDA) was chosen as the method of classifying papers into comparable topics [38–42]. A STATA syntax was used to assign papers to 10 major topics. The data were exported into Microsoft Excel for reading titles and abstracts. Two researchers independently reviewed the titles and abstracts of most cited papers within each research topic to assign the labels correctly for each topic. After that, an expert would discuss with two researchers to unify the name of topics. Besides providing the total number and percentage of research studies by each topic, we ranked the research interests of these research topics based on the total number of its publications in the past five years. Table 1 identifies how the inputs and outputs of each of these analytical methods.


