I thought the tide had shifted towards active voice as well. However, a year or two ago I looked it up and some well known journals still preferred passive voice.

Here's an interesting recent paper that found that roughly 30% of papers are still in the passive voice (admittedly, I was expecting a higher percentage).