Since we're bringing up rating and grading, and I'm totally not slacking off when I'm supposed to be preparing a class -
Teachers distinguish roughly between formative and summative assessment. Summative assessment is the kind of thing that goes on your transcript and measures your performance looking back ("Inspection is after the fact, after the parts are produced.") It does have value, but its value is not directly in improving the process or the student's learning.
Formative assessment on the other hand, although it might use measurements or grades, doesn't feed these into metrics or transcripts, and definitely not into bonuses or penalties. Any numbers produced by a formative process are not outputs, but inputs to a conversation that ideally looks at more than just the numbers. That's what you want to improve the process that produces the parts (or the learning) in the first place.
It sounds to me like some of the examples in the page, such as the student who was sent to remedial training after receiving two test scores below average in a row, are a kind of cargo-cult Deming process: you might have measurements and control charts but you're not actually using them the way they were intended.
This is also, by the way, why I am extremely skeptical to say the least of any attempts to "use technology to make education more data-driven". It's not that the idea couldn't possibly work, but I have yet to see what I'd consider a good implementation.
Teachers distinguish roughly between formative and summative assessment. Summative assessment is the kind of thing that goes on your transcript and measures your performance looking back ("Inspection is after the fact, after the parts are produced.") It does have value, but its value is not directly in improving the process or the student's learning.
Formative assessment on the other hand, although it might use measurements or grades, doesn't feed these into metrics or transcripts, and definitely not into bonuses or penalties. Any numbers produced by a formative process are not outputs, but inputs to a conversation that ideally looks at more than just the numbers. That's what you want to improve the process that produces the parts (or the learning) in the first place.
It sounds to me like some of the examples in the page, such as the student who was sent to remedial training after receiving two test scores below average in a row, are a kind of cargo-cult Deming process: you might have measurements and control charts but you're not actually using them the way they were intended.
This is also, by the way, why I am extremely skeptical to say the least of any attempts to "use technology to make education more data-driven". It's not that the idea couldn't possibly work, but I have yet to see what I'd consider a good implementation.