Using a language that supports the creation of strongly typed source code is half the story. Developers have to make extensive use of the type functionality to get the benefits.
Comparing just the built-in type checking rules (e.g., not allowed to add a string to an integer), then languages are very similar, e.g., C++ vs Ada
Being effective requires monitoring performance; how well are things going before/after a change. An understanding of how things work is needed to guide new ideas; randomly blundering around in is very ineffective.
What is the most effective way of doing software engineering?
There is lots of folklore and existing practice, but no evidence that any of it is the best way of doing things.
There is an analysis of all the available public software engineering data (as of 2020), but the 600+ datasets are not enough to build reliable models.
What is needed is more data. If anybody knows of any public software engineering data, please let me know. I’m always up for analyzing software data. I know there is lots of code on Github, and there are lots of animals in the Jungle; this is not data. The hard to find data is human related, e.g., task estimates and schedules.