Fig 1.
Time to complete the survey.
Fig 2.
Frequency of individual scores: Blue line includes all responses, N = 6604, and an average score, S, of 48.6; bars represent retained responses, N = 5781 and an average score of 53.7; red line is the best fit Gaussian distribution, R2 = 0.985.
Table 1.
Participation rates for each category j, with a total number of participants Nj, and total score, Sj as a function of scientific field and category.
We grouped the 250 WoS sub-categories into 21 categories and then combined these into 4 fields based on the total score (Sj) and a t-test comparing si.
Fig 3.
All fields Likert chart: 5500 respondents.
The number of responses less than zero on the x-axis include almost never (ω0), sometimes (ω1) and half of the often (ω2) responses. The number of responses greater than zero include almost always (ω4), usually (ω3), and half of the often responses. The mean scores on the y-axis are and varies from a low of 0.7 (troubleshooting) to a high of 3.7 (drafting).
Fig 4.
Mean scores, si, according to scientific category.
We grouped 5363 respondents according to the Clarivate Analytics science category definitions [31] and then calculated a mean score per question for each category based on the total number of respondents per category, ni: .
Fig 5.
Mean scores according to scientific field.
We compared the aggregate category scores, S, and assigned them to one of four fields based on the Student’s t-test. Pure and applied sciences includes physics, chemistry, materials, and engineering (S = 59). Natural sciences includes biology, geosciences, medicine, computer science, agricultural sciences, environment, statistics, and archeology (S = 54). Many humanities categories made up the third field (S = 48): psychology, management, mathematics, sociology, literature, education, and health. The final field (S = 41) comprised philosophy, political science, and law. The error bars depict 95% confidence intervals, ni is the number of responses per category, and . (Symbol sizes are larger than most error bars).
Fig 6.
Scores for each of the questions with respect to career experience.
The data set includes responses from pure and applied sciences and natural sciences and excludes those from Asia and Africa. The questions were organized from lowest to highest score based on the answers of the professionals with 5 years of experience where ni is the number of responses for each of the 4 categories of experience and .
Fig 7.
Number of respondents from each country and grouped in regions.
Fig 8.
Activity scores as a function of region.
Researchers from 115 countries answered the survey and we grouped them into regions based on t-tests comparing the mean score () of the country and region. The Latin region includes South American countries, Portugal, Spain, Italy, France, Greece, and Cyprus. Anglophones include the United States and many Commonwealth countries. Slavic nations comprise Eastern Europe and former Soviet states. India, Pakistan, Indonesia, Singapore, and Bangladesh are part of the South Asia region. The Germanic states include European countries north of France/Italy. Middle East and Africa (MEA) is the most diverse grouping and includes Iran, Turkey, Arabic-speaking nations and sub-Saharan Africa (some of which could have been included among the Commonwealth nations). China, Japan, South Korea, Taiwan, Vietnam comprise the Far East Asia region (FEA).
Fig 9.
Medical sciences Likert chart: 720 respondents. The number of responses less than zero on the x-axis include almost never (ω0), sometimes (ω1) and half of the often (ω2) responses. The number of responses greater than zero include almost always (ω4), usually (ω3), and half of the often responses. The mean scores on the y-axis are . si varies from a low of 0.7 (troubleshooting) to a high of 3.8 (drafting).