Should citations be field-normalized in evaluative bibliometrics? An empirical analysis based on propensity score matching

•We investigated the question how strong fields influence citation counts of papers.•Can field differences be traced back to factors possibly influencing citations (FICs)?•We considered 13 variables as FICs, e.g., number of pages, co-authors, and cited references.•We applied propensity score matching as statistical method to make causal statements.•The results revealed that field differences did not completely vanish but were dramatically reduced.