Page 13 - The lraternational Journal of the Royal Society of Thailand.indd

The International Journal of the Royal Society of Thailand
Volume XV-2023

If the number of these frequent co-authors exceeded 10, we included only the top 10
frequent co-authors. In addition, frequent co-authors who do not affiliate with an
organization in Thailand, as well as those who have co-authored fewer than 3 articles,

were excluded. Subsequently, publication records of the remaining frequent co-authors
of the Index Researcher IR_01, regarded as Researchers in Circle 1 (RC_1), were retrieved
from the database for further analyses. The snowball sampling process was iterated

for all members of RC_1 to incorporate their frequent co-authors into Circle 2 (RC_2),
and this process continued until we reached Researchers in Circle 4 (RC_4).

Characteristics of co-authorship networks
Publication records of each included individual researcher were employed

to establish a weighted, ego-centric co-authorship network, using igraph package
(Csardi & Nepusz, 2006). Given the significant impact of paper mills on the academic

community in recent years (Abalkina, 2022), we restricted our analysis to publication
records from the past five years, spanning from January 31, 2018, to January 31, 2023.
Additionally, we restricted our analysis to publications with fewer than 25 co-authors

to emphasize co-authorships characterized by relatively equal contributions, as opposed
to those with a larger number of co-authors.

In the co-authorship network of each researcher (ego), nodes represented
co-authors (alters) of the ego. Edges in the network were non-directional and

indicated that two researchers had collaborated on an article, with the weight of the
edges determined by the number of articles they co-authored. Publication record

profiles of each researcher were subsequently described using three following
categories: 1) Researcher (ego) characteristics; 2) Co-author (alters) characteristics; and
3) Ego-centric co-authorship network characteristics.

Firstly, researcher characteristics encompassed the following variables: 1) Time
since the first publication (in years), reflecting their experience in their research career;
2) Average annual publication count over the past 5 years; 3) Maximum annual

publication count over the past 5 years; 4) Proportions of published articles categorized
as review articles, systematic review articles, or meta-analysis articles; and 5) Diversity
of ANZSRC 2020 research domains, calculated using Shannon’s index of diversity

(entropy) (Shannon, 1948).

The characteristics of co-authors were described as proportions of the geographical
subregions of their respective countries of affiliation, calculated as a ratio to the total

number of co-authors. The geographical subregions were classified following the M49

Phrutsamon Wongnak et al. 5