Skip to Main Content
Table 1.

Results of logistic regression for UPRL, of FP-funded publications with at least one UK-based author, 2008–2016 (source: Scopus and Overton)

VariableLogistic regression for variables in logarithmic form and in level
Log-transformed (Model 1)Level (Model 2)Level – for one standard dev. (Model 3)
CoefficientsOdds ratioCoefficientsOdds ratioCoefficientsOdds ratioΔp (baseline = 10%)
Normalized citation score 0.960*** (0.035) 2.613 0.086*** (0.006) 1.09 0.613*** (0.043) 1.846 7.0% 
Normalized CiteScore 0.024 (0.063) 1.024 0.153*** (0.021) 1.166 0.207*** (0.029) 1.23 2.0% 
Document type (Review = 1) 0.727*** (0.101) 2.07 0.507*** (0.096) 1.661 0.507*** (0.096) 1.661 5.6% 
Multidisciplinary (DDA) 0.234*** (0.058) 1.264 0.109*** (0.026) 1.115 0.129*** (0.031) 1.138 1.2% 
Interdisciplinarity (DDR) 0.017 (0.125) 1.017 0.07 (0.089) 1.073 0.032 (0.041) 1.033 0.3% 
Number of authors −0.111** (0.048) 0.895 −0.006*** (0.001) 0.994 −0.780*** (0.176) 0.458 −5.2% 
Number of countries 0.163*** (0.063) 1.178 0.074*** (0.011) 1.077 0.279*** (0.040) 1.322 2.8% 
Proportion of female authors 0.300** (0.152) 1.35 0.152 (0.143) 1.164 0.033 (0.031) 1.033 0.3% 
Avg. number of papers per author 0.107*** (0.040) 1.113 0.003*** (0.001) 1.003 0.115*** (0.039) 1.122 1.1% 
Observations 37,897   37,897   37,897     
Log Likelihood −3,446.58   −3,754.04   −3,754.04     
Wald Test (df = 9) 1,046.660***   528.910***   528.910***     
LR Test (df = 9) 1,345.617***   730.702***   730.702***     
Score (Logrank) Test (df = 9) 1,316.022***   766.147***   766.147***     
VariableLogistic regression for variables in logarithmic form and in level
Log-transformed (Model 1)Level (Model 2)Level – for one standard dev. (Model 3)
CoefficientsOdds ratioCoefficientsOdds ratioCoefficientsOdds ratioΔp (baseline = 10%)
Normalized citation score 0.960*** (0.035) 2.613 0.086*** (0.006) 1.09 0.613*** (0.043) 1.846 7.0% 
Normalized CiteScore 0.024 (0.063) 1.024 0.153*** (0.021) 1.166 0.207*** (0.029) 1.23 2.0% 
Document type (Review = 1) 0.727*** (0.101) 2.07 0.507*** (0.096) 1.661 0.507*** (0.096) 1.661 5.6% 
Multidisciplinary (DDA) 0.234*** (0.058) 1.264 0.109*** (0.026) 1.115 0.129*** (0.031) 1.138 1.2% 
Interdisciplinarity (DDR) 0.017 (0.125) 1.017 0.07 (0.089) 1.073 0.032 (0.041) 1.033 0.3% 
Number of authors −0.111** (0.048) 0.895 −0.006*** (0.001) 0.994 −0.780*** (0.176) 0.458 −5.2% 
Number of countries 0.163*** (0.063) 1.178 0.074*** (0.011) 1.077 0.279*** (0.040) 1.322 2.8% 
Proportion of female authors 0.300** (0.152) 1.35 0.152 (0.143) 1.164 0.033 (0.031) 1.033 0.3% 
Avg. number of papers per author 0.107*** (0.040) 1.113 0.003*** (0.001) 1.003 0.115*** (0.039) 1.122 1.1% 
Observations 37,897   37,897   37,897     
Log Likelihood −3,446.58   −3,754.04   −3,754.04     
Wald Test (df = 9) 1,046.660***   528.910***   528.910***     
LR Test (df = 9) 1,345.617***   730.702***   730.702***     
Score (Logrank) Test (df = 9) 1,316.022***   766.147***   766.147***     

Note: Binary logarithms of normalized citation counts, normalized CiteScore, DDA, DDR, number of authors/countries, and average number of papers per author were used in Model 1. Therefore, the odds ratio of these coefficients refers to the variation in odds associated with a twofold change in the explanatory variables. *p < 0.1; **p < 0.05; ***p < 0.01.

Close Modal

or Create an Account

Close Modal
Close Modal