When discussing formative assessment and my current standards-based grading system with colleagues, various aspects of it seem like a lot of "work" (such as inputting multiple scores into the grade book rather than just a single summative assessment score) while other parts are viewed as downright controversial. 

In Revisiting Professional Learning Communities at Work, DuFour, et. al (2008) suggest three components of formative assessment:

  1. The assessment is used to identify students who are experiencing difficulty in their learning. 
  2. A system of intervention is in place to ensure students experiencing difficulty devote additional time to and receive additional support for their learning.
  3. Those students are provided another opportunity to demonstrate their learning and are not penalized for their earlier difficulty. (emphasis mine, pp. 216-217)
Allowing new evidence of learning to replace old evidence is such a hard sell.  I hear responses such as Won't students give a mediocre effort the first time?  We're not teaching responsibility!  This isn't how college works.  I've used this example in a previous post and it's also the one I use to illustrate the point that new evidence of learning should replace old evidence:
Consider the following example. Assume that homework is graded on completion and quizzes/tests on content mastery.

Student A: Homework: 50% Quiz: 60% Test: 100%
Student B: Homework: 100% Quiz 100% Test: 100%

Student A did not understand the concepts and therefore did not complete the homework. Somewhere between the "quiz" and the "test" Student A came in for extra help and finally "understood" the concept which explains his/her sudden improvement on the "test."

In the traditional grading system, which student earns a better grade? Student B, of course. A traditional points system penalizes "later learners." On the "test," both students demonstrated the same level of understanding, but Student A is penalized for initially struggling. Do we have a realistic expectation that students will "get it" the first day we teach concepts to them? If so, then why not have daily tests?
DuFour, et. al go on to explain this point concisely in their book: 
"Our position has been challenged in several ways. Some have argued  students should not be given a second opportunity to learn, or, at the very least, their initial failure should be included in calculating the grade. They claim it would be unfair to allow low-performing students the opportunity to earn a grade similar to those of students who were proficient on the initial assessment. Our response is that every school mission statement we have read asserts the school is committed to helping all students learn. We have yet to find a mission statement that says, “They must all learn fast or the first time we teach it.” If some students must work longer and harder to succeed, but they become proficient, their grade should reflect their ultimate proficiency, not their early difficulty." (p. 219)
I am becoming increasingly convinced that any classroom claiming to involve formative assessment or "assessment for learning" must allow new evidence to replace the old.  It just makes sense.