How can we properly assess student learning?

This is a hard question to answer and, quite frankly, I don’t have the solution.  I am of the opinion that writing assessments are they are the best way to assess learning.  For teachers, however, they may be the hardest to manage.  They definitely take a lot longer to grade than running a bubble sheet through a Scantron machine to spit out a score.  But what about students who are anxious about writing essays, who don’t write particularly well, or need some form of scaffolding to prompt their memories or at least offer problem solving options (i.e. multiple-choice, where some students can minimize the risk of a wrong answer by properly discarding incorrect choices)?

Can one assessment format work well for everyone?  Are performance-based exams (playing a musical instrument, participation in group work, more complex, structured assignments such as term papers, etc.) or personal communication assessments (oral exams, journals, blogging, etc.) better?  I’m not sure.  The most logical answer (to me) is that each has its place.  That might be OK in the classroom, but what about standardized readiness assessments?  I’m not sure.  What about teachers or test makers who design their assessments to try to “out-guess” students?  Should the goal of an exam be to confuse or introduce roadblocks to understanding?  I think not; however, I believe this happens.  A lot.

The Smarter Balanced Assessment Consortium has been tasked by the U.S. Department of Education to accomplish the goal of a “better” test – a “high-quality assessment system”.  This exam will be facilitated by computer and is a “learning” exam (computer adaptive technology) that adjusts its future questions based on the response of past questions (incorrect answers get easier follow-up questions, correct answers get harder follow-up questions).  In Michigan, the transition will begin for the 2014-2015 school year; however, all schools are not yet able to meet the technology requirements and therefore a paper and pencil option will still be available.  Will students who take the computer adaptive exam score similarly to those who do not?  Would these students have scored similarly on the paper and pencil exam?  Can two different test formats be properly aggregated, compared to, or ranked against each other?

One area I contend that teachers and test makers can work on is to proactively build better tests.  The goal should be to write great questions - questions that gauge understanding and allow educators to see a student’s thinking.  If the questions, in any format, are easily understandable and allow the test taker a true opportunity to show his or her knowledge, not contain it within the test maker’s frame of reference, we will be closer to answering the question of assessment value?