Today, the role of science and science production in the growth and development of countries is obvious. Scholarly publications are crucial for higher learning institutions, as they lead to the creation of new knowledge, enhance reputation, stimulate modernization and innovation, improve the quality of academic staff, and boost the institution’s economic status (Dhillon et al., 2015; Wahid et al., 2022). Thus, most researchers are striving to enhance their scientific output to increase their national and international rankings. Science is driven primarily by the efforts of a small number of researchers who publish the majority of scientific research and receive a large number of citations (Parker et al., 2013). Previous studies have indicated that many variables influence the number of publications by a scholar. The impacts of researchers’ age, gender, professional position, field of study, and geographic differences on publication rate have been studied to some extent. For instance, gender-related studies have shown that women tend to publish fewer publications than men, and geographic differences are less important (Abramo et al., 2009; Kyvik & Teigen, 1996; Xie & Shauman, 1998). It is reported that publishing productivity increases with age and tends to decline at a certain age, usually in the early forties, even though the productivity peak varies between fields (Cole, 1979; Kyvik, 1990). However, the effect of age was not always significant. For instance, a lack of notable correlation between age and research productivity has been documented (Ramsden, 1994; Williams, 2001). Some studies have examined one or a few of the aforementioned factors within a single country, institute, or field of science (Bonaccorsi & Daraio, 2003; Lima et al., 2015; Wallner et al., 2003; Way et al., 2017). For instance, economists’ publication productivity in top journals decreases significantly with age, while acceptance rates remain consistent regardless of age (Oster & Hamermesh, 1998). Other parameters that may affect research productivity are availability of funding for research, teaching responsibilities, access to equipment, assistance from research staff, workload expectations, departmental culture and work environment, organizational setting, as well as individual talent and dedication (Dundar & Lewis, 1998; Kyvik, 1993; Ramesh Babu & Singh, 1998). While most previous studies have examined the effect of gender, age, field of study, and geographical differences individually (Puuska, 2010; Shin & Cummings, 2010), to the best of our knowledge, no work so far has studied the mentioned variables simultaneously.

In addition to the above-mentioned variables, it appears that one’s position (e.g., being a university chancellor) is another influential variable on the number of publications. We know that top scholars may be recommended to lead prestigious universities (Goodall, 2009). It is also appreciated that holding the position of chancellor at a university demands a significant amount of time and energy for executive duties; therefore, less time and energy will be available to conduct research. It is therefore expected that the number of articles will decrease after a while. On the other hand, the chancellor holds significant authority and influence over all university subunits, enhancing networking both within and outside the university; consequently, an increase in the number of articles is expected. Although the significance of power is widely acknowledged in politics, economics, and psychology, its role in academia has received relatively little focus. It is believed that academic merit dictates resource distribution. Nevertheless, in power-oriented societies, the possible effects of power dynamics on academic institutions cannot be ignored (Jia et al., 2019; Schilke et al., 2015). This study is the first to evaluate the relationship between university chancellorship and the number of publications.

Using the QS world university rankings database in 2023, the top 1,000 universities were coded. Subsequently, among these universities, chancellors who met the following admission requirements were included in the study until the required sample size was reached:

• Duration of chancellorship: minimum 3 years

• Beginning of the chancellorship: Between 2000 and 2020

• Having author profile in Scopus Database (www.scopus.com)

In this study, consistent with the educational system of each country and for easier access to information, the terms chancellor, president, and rector were used interchangeably.

As there was no similar study to estimate the proper sample size, the effect size of the studied variables (the difference in the number of articles before and after the beginning of the chancellorship) was investigated during a pilot study. Based on the pilot study findings, the sample size was estimated at 196 using G-Power 3.0 software, considering the error type I (alpha) and type II (beta) equal to 0.05 (test power of 0.95). Thus, the sample of this study consisted of 196 chancellors drawn randomly (using Excel software) from the world’s top 1,000 universities.

This study used an observational, analytical, and longitudinal design. The number of articles was collected using the Scopus database for a six-year period (three years before and three years after the beginning of the chancellorship). The Scopus database was used to collect the data because it offers a wider range of scholarly information, facilitates detailed searches about authors, is updated daily, provides access to reliable data, metrics, and analytical tools, and provides access to researcher profiles in which lists of publications are displayed.

A checklist containing variables such as gender, age, continent, and being active or non-active was employed to collect the data. Active professors were those who had published at least one article each year in the three-year period before beginning their chancellorship. The remaining chancellors were defined as inactive. The mean number of articles during the period of three years before to three years after the beginning of the chancellorship was compared using the Friedman test.

Table 1 shows the demographic variables of the selected university chancellors, which includes the variables of age, gender, activity, continent, and activity.

The age of the university chancellors varied between 40 and 86 years, with a mean and standard deviation of 60.66 and 9.06, respectively. As shown in Table 1, the highest and lowest frequencies of age groups were associated with the 51-60 and 81-90 groups, respectively. Most university chancellors were male (88.3%). As for continents, Asia and Oceania showed the highest and lowest frequencies, respectively.

Figure 1 shows the trend in the number of publications as a function of chancellorship. As Figure 1 shows, there was generally a descending slope in the publication number after the chancellorship beginning (CB) year. According to the Friedman test, the means of the number of articles published by the chancellors at each time point were statistically different (p<0.001). The pairwise comparison of these means using Bonferroni correction showed statistically significant differences between the time points before and after CB (see Supplementary Table S1). Furthermore, a paired t-test showed that the difference in the mean number of publications during the three years before CB and two years after CB was statistically significant (t=3.57, df=195, p<.001).

To further study the relationship between chancellorship and the number of publications, the chancellors were categorized based on their activity, gender, age, and continent. Then, their publication numbers were compared in three years before and after the chancellorship, and the results are given in Figures 2-5. Figure 2 shows the trends in the publication number of active and non-active chancellors over the six time points of the study. As evidenced, no obvious fluctuation is observed in the trend pertaining to non-active chancellors, while among active chancellors, a reduction in their publication number is visible. The number of articles published by active and non-active chancellors in each year was also compared statistically. According to the Friedman test, the means of the number of published articles by non-active chancellors at each time point were not significantly different (p>0.196), while among active chancellors, these means showed a significant difference (p<0.001). Table S2 (see Supplementary) presents the results of pairwise comparison of the means of the number of published articles by the active chancellors in each year using Bonferroni correction. Evidently, among active chancellors, the means of the number of publications in the first and second years after CB were statistically significantly lower than the means of the publication number in other years.

Figure 3 illustrates the relationship between chancellorship and the publication numbers of male and female chancellors. Considering the details, the means of the number of publications at each study time point among male chancellors were statistically significantly different (p<0.001), while these differences were not significant among female chancellors (p=0.242). Table S3 (see Supplementary) details the findings of pairwise comparison of the number of published articles in male chancellors using the Bonferroni correction method. Of the 15 comparisons, four were significant in male chancellors.

Figure 4 shows the number of publications of chancellors as a function of their age. The Figure shows that the number of publications in chancellors decreased significantly in both age groups (i.e., younger and older than 60). Pairwise comparisons indicated one significant difference (p=0.026) for chancellors younger than 60 years (one year before CB versus two years after CB) and two significant differences for chancellors older than 60 years (one year before CB versus two years after CB and one year before CB versus one year after CB) (p<0.05).

Figure 5 compares the mean number of published articles by chancellors based on their continent. Using the Friedman test, the only significant difference was observed in North America’s chancellors between three years before CB versus one year after CB (p=0.025). Comparisons for the remaining continents were not statistically significant (p>0.05). The results of multiple comparisons using the Bonferroni correction to compare the data for North America showed one significant difference (see Table S5 in the Supplementary).

Finally, in a separate study, the data of the two chancellors with the highest number of publications were removed from the list of 196 chancellors (i.e., initial list of chancellors). All the above-mentioned comparisons were then repeated to remove the potential impact of outliers. The results remained consistent, which shows that the data of chancellors with very high publications do not influence our findings (details not given).

This study evaluated the number of articles of chancellors of the world’s top universities in the period of three years before and three years after the beginning of their chancellorship. Several factors are associated with research productivity. These factors include socio-demographic and career factors, as well as self-knowledge, social knowledge, behaviors, and environmental constructs (Blackburn & Lawrence, 1995). In a study of science faculty members, three categories of correlates of research productivity were identified: individual characteristics (such as psychological characteristics, work habits, and demographics), work environmental factors, and reinforcing feedback (colleagues and mentorship) (Fox, 1983). The results of our study indicate a decline in the number of publications by chancellors after they began their tenure. It is arguable that being engaged in executive tasks leads to an essential reduction in the time required for designing and conducting research. In other words, the assumption that increasing the reputation and available resources (Abramo et al., 2016), which is associated with beginning the chancellorship, may enhance the number of articles, does not appear to work for chancellors. In a previous publication, the impact of deanship on academic publication among economics school deans in Chinese universities was examined. The findings indicated that deanship leads to an increase of 0.7 publications per year. The increase was primarily driven by collaborations with colleagues within the same institution, often on topics outside the deans’ previous research areas. The effect of deanship on publication rates was less significant in top universities and leading journals. The impact was also less pronounced among deans with strong publication records prior to their appointments. The results suggest that the increase in publications is more attributable to resource allocation influenced by deans’ power rather than their abilities or reputations (Jia et al., 2019).

Our findings indicate that the reduction in the number of publications is only significant in active chancellors. It is arguable that the very low number of publications of non-active chancellors (commonly zero or one each year) is not affected by the limited time available for conducting the research. In active chancellors, reduced free time leads to a reduced number of publications. Additionally, the number of non-active chancellors (i.e., nearly half the total number of chancellors under study) does not appear to agree with the conclusion of Goodall, saying that highly ranked universities are led by top researchers (Goodall, 2009).

Investigating the effect of the beginning of chancellorship as a function of gender showed that this effect is only apparent in male chancellors. Previously, it has been reported that regardless of admin position, the publication number in women increases continuously by age, while it is curvilinear in men (i.e., reaches a plateau at 60-64 and decreases afterwards) (Rørstad & Aksnes, 2015). It is therefore arguable that in addition to age, administrative positions have the potential to significantly affect the academic performance of men. Furthermore, in general, men have a higher publication number than women up to the age of 55-59 years.

Furthermore, although previous literature shows the importance of age on the productivity of academics (Gingras et al., 2008), our study indicated that the two age categories (i.e., younger and older than 60) are more or less affected by chancellorship. Previously, in a comparative analysis, it was found that academic performance is more dependent on the work environment than age. Higher positions in the hierarchy are effective in improving publication output, owing to better access to financial resources (Knorr et al., 1979).

Geographically, the chancellors in this study indicated a decrease in the number of publications in only one group (North America), having mentioned that chancellors in Asia and Europe indicated slight decreases, which were not significant. This could be due to the fact that what a chancellor does in different universities is different. Therefore, no common rule can be devised for different countries.

Implications for consideration:

1. This study suggests that assuming a chancellorship at a top-ranked university may negatively impact academic output in terms of published articles. This raises questions regarding the potential reasons for this decline.

2. This study suggests that taking on executive responsibilities, such as serving as a university chancellor, may decrease the time available for research activities, leading to a reduction in publication output. This finding highlights the challenge of balancing administrative duties with academic research.

3. This study indicates that male chancellors experience a significant decrease in publication numbers after assuming their roles, while the effect is not as prominent in female chancellors. This suggests that administrative positions may have different impacts on the academic performance of men and women, possibly influencing publication patterns based on gender.

4. The research identifies regional differences in publication output among university chancellors, with notable decreases observed in North America but not significant changes in Asia and Europe. This suggests that the effects of chancellorship on research productivity may vary based on the geographic location and institutional context.

5. The study’s findings offer valuable insights for policymakers regarding the impact of administrative roles, such as university chancellorship, on the academic output of university academics. Understanding these implications can help develop strategies to support and enhance research productivity in academic leadership.

6. These outcomes also indicate that further research is needed to understand and address the potential challenges faced by academic leaders in maintaining their own scholarly productivity while fulfilling their administrative duties. This has implications for universities and academic institutions in terms of supporting their leaders in effectively managing these dual responsibilities.

7. A limitation of our study was that in some universities, the chancellor does not govern the university; instead, the vice-chancellor handles administrative duties.

Amir Amani (amani76@gmail.com): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Project administration (Equal), Writing - original draft (Lead).

Mohammad Reza Armat (rezaarmat@yahoo.com): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Writing - review & editing (Equal).

Samaneh Mollazadeh (samanehmollazadeh@yahoo.com): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Writing - review & editing (Equal).

Reza Salarinia (rezasalarinia@gmail.com): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Writing - review & editing (Equal).

Mitra Salehi (misssalehi@yahoo.com): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Writing - review & editing (Equal).

Sonia Fathi-karkan (soniafathi92@gmail.com): Data curation (Equal), Investigation (Equal), Methodology (Equal), Writing - review & editing (Equal).

Akbar Solati (akbarsolati90@gmail.com; ORCID: 0009-0001-1508-0299): Conceptualization (Equal), Data curation (Equal), Formal analysis (Equal), Funding acquisition (Equal), Investigation (Equal), Methodology (Equal), Project administration (Equal), Supervision (Equal), Validation (Equal), Visualization (Equal), Writing - review & editing (Equal).

The research was funded by the North Khorasan University of Medical Sciences, grant No. 4010231.

-pdf file

The data can be accessed at: https://doi.org/10.57760/sciencedb.25016