I previously discussed why “soft” research (qualitative research) is so much harder than “hard” research (quantitative research). The main reason is that there is less and later feedback, which can be incredibly frustrating for the impatient researcher. A similar argument applies to teaching “soft” skills, which is much harder than teaching “hard” skills.
Examples of hard skills are math or programming, examples of soft skills are teamwork or coaching other people. When teaching math or programming, feedback comes easy: Either the results are correct or not. And, it is comparatively easy to find out whether some student homework is correct or not. Case closed. People call this hard, because feedback is fast and rabid, more specifically: Not pleasant if you get it wrong because there is no wiggle space. You can’t argue much if the math doesn’t work out or the program produces bugs.
In many people’s eyes, this is different for soft skills. They are easy, because results aren’t black or white, but shades of gray: You can be good or bad, but you will always get at least some points. Maybe.
I think that many professors discount soft skills, because they have had no training in them and don’t know how to teach them. Specifically, grading soft skills is much harder than grading math or programming. As always, you have to set up an evaluation model, and then you have to carry it through. With qualitative data as input, grading becomes more time intensive and more prone to complaints from students who are angling for better grades.
Soft skills are as important to a student’s professional success as are hard skills. Pulling your feet away because you don’t know how to teach them is an easy but in my book not acceptable way out. I have been othered by my colleagues as a sociologists many times for teaching both hard and soft skills (and not just hard skills). But as a software engineering researcher and teacher I can’t do differently: Communication and coordination of people is a critical part of software engineering.