I remember reading an old research paper (pre-minecraft) that tracked long-term effects of kid-friendly languages like logo and scratch-like languages vs textua

I remember reading an old research paper (pre-minecraft) that tracked long-term effects of kid-friendly languages like logo and scratch-like languages vs textual-based programming like basic. It found that the former got more kids engaged but there was minimal lasting impact (none of those tracked became programmers or stated any interest in learning more as adults). The later got fewer kids interested initially but those who did get interested had a high chance of carrying those skills into programming careers. The results would probably be very different now, but that paper really colored my thinking when I first read it.