Using quality assessment criteria for valid assessment results to guide student achievement goals
Sound assessment results are crucial when teachers are making many important decisions about their goals and teaching based upon them. This session showed participants how to review and revise existing assessments for quality based on key criteria.
A balanced assessment system has 3 types of assessments:
- Large Scale (Assessment of) - these are summative, norm-referenced, aptitude, and achievement tests. The ACT and the TABE would be considered large scale. Essential Question: What have students already learned?
- Mid-Scale (Assessment for) - these are formative, criterion-referenced, often teacher- or district- made, and achievement tests. Essential Question: How can we help students learn more?
- Small-Scale (Assessment for) - these include questioning, day by day, minute by minute, and achievement tests. Most of our time and energy should go to these. Essential question: How can we help students learn more?
There are two types of tests: norm-referenced tests and criterion-referenced tests.
- Criteria-referenced: state tests are criteria-referenced, meaning that they measure kids against criteria. It is possible for all students to be successful on these tests.
- Norm-referenced: the ACT is norm-referenced, meaning that it is designed with a score spread in mind. It is not possible for all students to be successful on this test - we can never meet proficiency in a norm-referenced exam. Therefore, it cannot be used to show growth.
When writing an assessment, check that...
- Assessments reflect the essential learnings and standards
- The students have had an opportunity to learn the content
- Assessments are as free from bias as possible
- Assessment levels are appropriate
- There is consistency in scoring
- The mastery levels are appropriate
*Proficiency Scales and Achievement Level Definitions for Cut Scores (p.229-230)
*Angoff Method (p.241-243)
My opinions:
These methods are time-consuming but seem to yield much more reliable multiple choice and short answer tests. Because I hardly use multiple choice tests and am trying to move away from a percentage-based grading scale, I probably will not be using these methods. However, this session would be of interest to teachers who largely rely on those types of tests to determine their students' grades.
This is why it ticks me off that the AP exams are norm-referenced (and tons of college courses, for that matter). Why do we REQUIRE that 50% of the kids taking the test do not get college credit? Or, for that matter, if all the teachers sucked one year and kids didn't have the knowledge/skills they should, why do half of them get college credit? What a stupid system...who came up with it?
ReplyDeleteThis session is linked to Guskey's session on Grading and Reporting. In the Breakout, Tammy actually broke down how to make multiple choice and short answer tests more reliable. Teachers should look into this individually depending on the kids of assessments they give.
ReplyDeleteI think that we can use the 6 quality criteria to evaluate each others' assessments. Right now we are supposed to submit our final exams in advance in the spirit of backwards design. But right now I don't think we're following through with the backwards design to make this practice meaningful. One way it can be more meaningful is if we spent time in departments to examine each others' finals and discuss these 6 criteria. Does the final reflect the essential learnings? How did the students get the opportunity to learn the content?
ReplyDelete