Count highly-cited papers instead of papers with h citations: use normalized citation counts and compare "like with like"!

Teixeira da Silva and Dobránszki (Scientometrics. 10.1007/s11192-018-2680-3, 2018) describe practical problems in using the h-index for the purpose of research evaluation. For example, they discuss the h-index differences among the bibliometric databases. In this Letter to the Editor, we argue for abstaining from using the h-index. One can use normalized indicators instead.