What Is Being Measured? (Validity Basics)
A teacher uses a reading comprehension test filled mostly with vocabulary definitions to decide whether students understand a science article.
What is content validity?
Students take the same test twice under similar conditions and receive nearly identical scores.
What is test-retest reliability?
An end-of-unit science test includes several topics that were never taught.
What is content underrepresentation?
One teacher scores essays harshly while another scores generously.
What is scorer inconsistency?
A teacher uses multiple measures (tests, projects, observations) before deciding if a student needs intervention.
What is strengthening validity through multiple measures?
A math benchmark accurately predicts which students will struggle on the state math test later in the year.
What is predictive validity?
A spelling test produces wildly different results depending on the day it’s given.
What is low reliability?
A teacher uses homework grades (with heavy parent involvement) to decide mastery of standards.
What is contamination of measurement?
A test is so short that one missed question dramatically changes a student’s score.
What is insufficient sampling of content?
A teacher avoids using a single quiz score to make high-stakes grading decisions.
What is responsible score interpretation?
Students’ writing scores increase after instruction, but the assessment mostly measures handwriting neatness instead of writing quality.
What is construct validity?
A benchmark assessment consistently ranks students in the same order, even if scores change slightly.
What is score consistency?
An assessment designed for 5th graders is used to evaluate 3rd graders’ performance.
What is inappropriate use of an assessment?
Students misunderstand directions, leading to random guessing.
What is error variance?
A teacher revises test questions after realizing students misinterpreted them.
What is improving validity through test revision?