Distributions of the number of researchers (r) and the number of publications (p) by academic discipline

The total number of Venezuelan researchers analyzed in this study is 6015. Of these, 2823 are men and 3192 are women, all registered in the PPI. Table 1 shows the distribution by gender and scientific field, as well as the distribution by gender and academic discipline for the number of papers published as reported in the CVs of the researchers. The data are organized in descending order of the total number of researchers in each field.

Table 1 Distributions of the presence of researchers in journals by academic discipline and gender. 

When the data shown in Table 1 are analyzed, one observes the global percentage of women is 6.13% higher than men. This data is in line with the incorporation of women in higher education in Venezuela, which moved from 43% in 1959 to 59% in 2005 (^{Delgado de Smith and Rojas, 2009}). This feminization of Venezuelan research, especially in the field of Life Sciences, has also been reported in a recent paper by ^{Caputo, Requena y Vargas (2012)}. In Spain, a similar shift has been reported, though the range is somewhat lower. In 1972, women university professors stood at 15% and by the 1990s they represented 29% of the body of university faculties (^{Pérez Sedeño et al., 2003}). Nonetheless, in terms of papers published, male researchers account for 8.35% higher global percentage. With regard to published papers by Spanish researchers, there appears to be more output by male researchers (^{Mauleón and Bordons, 2006}). When each area is analyzed, this percentage varies from field to field; the category Other is excluded from the analysis because it groups fields with few researchers and papers published. For example, in disciplines of Medical Sciences (Med Sci), Pediatrics (Ped), Economy (Econ), Linguistics (Ling), Law (Law), Psychology (Psych) and Ethics (Ethics), the prevalence of women tops 60%. Nonetheless, when these percentages are compared to papers published by women in these fields, we find they are inferior to said prevalence, except for the field of Psychology, where the two percentages are nearly the same, i.e., 74.5% and 74.28%, respectively.

When the same comparison is made with regard to male researchers, the areas exhibiting percentages higher than 60% are Physics (Phys), Mathematics (Math), Earth Science (Earth Sci), Philosophy (Philo), and Astronomy and Astrophysics (Astro). When these values are compared to publication rates, we find these rates to be higher, except for the case of Philosophy, which came in slightly lower at 56.32%.

Figure 1 is a box-plot diagram of the data appearing in Table 1. One can observe that the average number of female researchers is somewhat higher than male counterparts, with Life Sciences and Medical Sciences exhibiting atypical numbers of female researchers. Moreover, the average number of papers published is lower among female researchers at 1885 versus the average for male counterparts at 2228. The number of paper published in the area of Life Sciences behaves distinctly for both men and women. Additionally, these exhibit atypical behavior in the fields of Technological Sciences (Tech Sci) and Medical Sciences.

Figure 1 Mosaic plot and box-plot of distribution of researches and papers published by field and gender 

Neither of the distributions exhibit significant differences with regard to gender of researchers, as shown by the variance analysis test for the number of researchers (F2,44=0.176, p-value: 0.6768) and for the number of papers published (F2,44=0.3099, p-value = 0.5805).

The analysis of data by gender of the number of researchers and papers published against the professional grade is shown in Table 2. Said table shows in general terms that women researchers have a larger presence at the lower professional grade levels, such as L_1 and el Candidate (Ca), representing 22.1% and 18.57%, respectively, of the total number researchers, while male researchers are represented in the higher levels (Up_L) at 16.37% of the total. Other research has obtained similar results, showing a presence of women researchers higher than that of male counterparts in the lowest professional grade levels, and lower presence in the higher levels (^{Abramo, D'Angelo y Caprasecca, 2009}; ^{Mauleón y Bordons, 2006}; ^{Mauleón, Bordons y Oppenheim, 2008}; ^{Caputo, Requena y Vargas, 2012}). In terms of scientific output of female researchers, this value came in higher at the highest professional levels (Up_L), at 22.6% of the papers published by women. Male researchers also boast higher scientific output at the highest level (Up_L), at 34.79% papers published by men, and at a proportion much higher than seen in women researchers.

Table 2 Number of researchers and papers published by gender and professional grade level 

An analysis of the presence of women researchers and their scientific output in each of the fields considered in terms of professional grade reveals that at the Candidate grade (Ca) women constitute majorities in 15 of the 22 fields, while men lead in the remaining seven, i.e., Technological Sciences (Tech Sci), Physics (Phys), Mathematics (Math), Earth Sciences (Earth Sci), Anthropology (Anthro), Astrophysics (Astro) and Ethics.

The number of publications by women at this level, however, is lower than the number of men in nine areas, including the seven already mentioned and in the Fine Arts and History (Arts).

At the next professional grade (L_1), there are fewer fields in which women are more numerous than men, since they are present in 13 of the 22 fields analyzed. In terms of papers published, the majority of fields in which women are more highly represented than men are also those in which they publish more numbers of papers, with the exception of Chemistry (Chem) where male researchers published 144 more papers than their female counterparts, despite being in a slim minority (two less researchers).

With regard to the highest professional grade (Up_L), the number of fields with greater numbers of women researchers falls to eight. When fields in which women researchers publish more papers than their male counterparts are analyzed, we find that seven of these fields coincide with those fields in which they are more highly represented. In Technological Sciences (Tech Sci), despite there being more women than men, the latter publish more papers than the former.

The gender and professional grade differences observed with regard to the number of researchers shown in Table 2 are not statistically significant, i.e., F_1,136= 0.4381, p-value = 0.5092) and (F_2,135= 1.236, p-value = 0.2939), respectively. In contrast, the differences in the number of publications are significant in term fo professional grade, i.e., F_2,135= 14.84, p-value < 0.05; but not for gender which came to F_1,136= 0.4894, p-value = 0.4854.

To determine whether the numbers of researchers and their papers published are spread equally across the diverse fields, a Gini index has been calculated for each of these distributions. Moreover, a Lorentz curve has been developed for each distribution. These results are shown in Table 3 and Figure 2, respectively, on the following page.

As can be observed, the concentration levels are not excessively high. The global behavior of both the number of researchers (0.4824) and scientific output (0.4916) are very similar. It is understood that there is greater homogeneity among researchers at the highest professional grade (Up_L), and among these, the distributions are somewhat more uniform for female researchers than is the case for their masculine counterparts. As for the other two professional grades considered (Ca and L_1), these show distributions are more concentrated for numbers of papers published than for the numbers of researchers.

Table 3 Gini Index: Distribution by area and gender. 

Figure 2 Lorenz curves de. Left.: by numbers of researchers. Right: by numbers of papers published. Men: Lines; Women: Dotted lines. 

Gender Parity Indexes (GPI)

Table 4 shows the gender parity indices calculated for each of the professional grades and fields. Taking all of the fields together (as is the case for the Other category, which is not weighed since several fields are grouped therein), we can see that for the number of researchers belonging to each field, eight of these fields (Pyschology (Psych, 2.929), Linguistics (Ling, 2.684), Law (Law, 2.200), Pediatrics (Ped, 2.187), Medical Sciences (Med Sci, 1.929), Economy (Econ, 1.547), Sociology (Sociol, 1.447) and Life Sciences (Life Sci, 1.255) female researchers are more prevalent than male researchers. In contrast, their values for professional output in these fields are lower than those for men across the board.Moreover, in two of these fields Sociology (Sociol, 0.966) and Life Sciences (Life Sci, 0.798), male researchers outperform female researchers, despite being in the minority.

Table 4 GPI (Gender Parity Index) of the number of researchers and number of papers published by field and PPI rank (Researcher Promotion Program). 

From the standpoint of professional grade parity, for three grades the results vary significantly from field to field (F_2,66= 3.234, p-value< 0.05). In general terms, it can be said that the results obtained in this research would show that the weight of women declines as their professional grade rises.

The average of the GPI value weighed as per number of researchers in each of the fields stands at 1.317 for the entire population, somewhat higher to the real proportion of women belonging to the system (3192/2823 = 1.131). This comes about because of the strong presence of women in the fields with higher numbers of human resources devoted to research, such as Life Science (Life Sci), Medical Sciences (Med Sci) and Pediatrics (Ped). The aggregate level of the presence of women declines with the professional grade, with the highest professional grade standing below the parity mark, since the values are above 1.80 at the lowest grade (Ca), and move to 1.47 at the middle grade (L_1), and to 0.83 at the highest grade (Up_L).

In terms of gender and field, a slightly different comportment is observed in the distribution of scientific output of men and women registered in PPI. The aggregate, as gauged by the weighted average of the number of publications, comes to a weighted GPI of 0.951, which signals, on one hand, global parity in the number of publications, but a loss of output efficiency among women, because it is lower than the value calculated for resources. As such, with regard to female output as a function of professional grade, one observes a relationship similar to that already obtained for the number of male researchers. This female output declines from its initial professional grade (Ca), where female output is 62.4% higher than their male counterparts, to much lower output values at the highest professional grades, where the balance tips significantly toward male output (0.734). At the intermediate professional grade (L_1), the output of women researchers begins to drop off, but in general terms remains above that exhibited by men by 37.6%. This diminishing output by female researchers (as they move up in professional grade) is seen across all fields, except once again in Technological Sciences (Tech Sci), where their output increases hand-in-hand with professional advancement. The differences observed in the parity indexes of the fields are not statistically attributable to professional grade of the researchers (F2,66=1.805, p-value = 0.1725).

Gender Success Rate (GSR)

Table 5 presents the Gender Success Rates for female researchers. Said table shows the ratio of output (number of papers published) to input (number of researchers) as each is gauged by its respective GPI value. The GSR value determines the observed gender performance against the expected gender performance gauged as a function of the proportion of female researchers in both outputs and inputs. In order to better understand this indicator, one must take into account that if in a given field women account for 80% against the population of men (GSR r=0.80), one would expect scientific output, as gauged by the number of papers published and reported in CVs to be similar. If the percentage of papers published by women against those published by men is higher than expected, then the GSR value would come in higher than 1.0, and we can safely speak of the greater success of women in the field where women outperform men.

Table 5 Output efficiency rate of male and female researchers by field and professional grade 

These conditions of success occur regardless of gender proportions in any given field; which is to say, there are fields with higher proportions of females who also publish more papers than the their male counterparts. These fields are marked with double asterisks (**) in Table 5. Despite this, male researchers are more efficient; for example, in the fields of Med Sci (GSR=0.753) and Ped (GSR=0.825). Likewise, there are fields with a higher proportion of female researchers, but which nonetheless exhibit fewer papers published than those published by men. These fields are marked with an asterisk (*) in Table 5. In this sense, there are fields such as Life Sciences (Life Sci), where despite the fact that women constitute the preponderance of researchers, they are less effective in terms of the number of papers published (GSR=0.636). This is the case across the three professional grades. As such, the GSR at the entry grade (Ca) of 0.694 and at the intermediate grade (L_1, of 0.847 is worthy of note; which is to say: even when women publish more papers than their male counterparts, they do so at a lower than expected proportion.

Additionally, for fields in which women are in the minority, we never observe the circumstance of women outperforming men in terms of papers published; but there are fields in which none of the three scenarios described above occur. These cases appear in Table 5 without asterisks. Interestingly, in this context, the field of Technological Sciences (Tech Sci), women are slightly more efficient than men (GSR=1.013), despite being in the minority (GPI-r=0.894) and less productive overall (GPI-p=0.906).

In the field of Physics (Phys), where women are in a small minority (GPI-r=0.21) and account for a low proportion of published papers (GPI-p=0.193), women exhibit gender efficiency near parity (GSR=0.919). In Chemistry (Chem) at the Candidate grade (Ca), women are more efficient than men; however, when analyzed as a whole, men turn out to be more efficient than women. In the social sciences, e.g., Economy (Econ, GSR=0.926) and Psychology (Psych, GSR=0.986), women exhibit values near gender parity, while in Sociology (Social) the index is much lower at (GSR=0.668).

Finally, the relational chart in Figure 3 presents the ten most productive fields for each of the three professional grades. The abscissa axis shows the GPI-r values and the ordinate axis shows the PPI-p values for the fields selected. Additionally, two axes have been marked at GPI=1 to differentiate those fields in which women are majority and in which their output is higher or lower than that of their male counterparts. The 45° line acts as a gender success boundary: the fields above it are those in which women exhibit greater success; and those below correspond to fields in which male researchers are most successful. Figure 3 shows women are more successful at the grade of Candidate (Ca). This is also true in the fields of Physics (Phys), Mathematics (Math), Chemistry (Chem) and Sociology (Sociol). Women at the grade of L_1 are more successful in the fields of Pediatrics (Ped), Agricultural Science (Agr Sci) and Earth Sciences (Earth Sci). With the exception of the field of Physics, which is shown practically atop the 45° boundary, there is no field in which women are more successful than men at the highest profession grade (Up_L).

Figure 3 Relational graph. Ten most productive areas (>3.000 papers published). Three professional grades.