Validity
The validity of the assessment is measured by considering if the task
• adequately reflects the domain of knowledge and skills of interest and
• if it can be used as the basis for the inference of proficiency
(Growing Success pg. 156)
***When an assessment is valid, the construct (ability, skill,
or knowledge) that it is designed to measure is the source of students’ score
on the assessment. In other words, it measures what it is designed to
measure.
Construct irrelevant variance (CIV) is the introduction
of extraneous, variables that affect assessment outcomes. The
meaningfulness and accuracy of examination results is adversely
affected, the legitimacy of decisions made upon results is affected,
and the validity is reduced.
Construct irrelevant factors can be:
·
Physical (hunger)
·
Environmental (noise in the hallway)
·
Social (trouble with friends)
·
Emotional (anxiety)
·
Bias (neater writing is marked
higher)
·
Systemic or Random
·
Can occur before, during or after an
assessment.
For the factors that we cannot control, we sometimes have to make decisions based on what is best for students. Maybe the assessment needs to be pushed off a day or a few hours until issues have resolved. Sometimes we need to rework the assignment so that these factors are diminished for students. Referring to a question as a "written response" and not an "essay", for example, could be enough to diminish a student's anxiety. Providing the materials, in class time and a variety of methods (as outlined in the UDL post) can eliminate some of these factors for a student.
Making sure the rubric is measuring what it is meant to measure will be the focus of the next blog post.