Chapter 3: Giving Meaning to Scores
Chapter 3: Giving Meaning to Scores
Nature of a Score
A score by itself lacks interpretative meaning.
For context, it is important to know whether a score is good or bad, e.g., a score of 44 could either be out of 44 (perfect) or 100 (poor).
Different tests measure different skills, e.g., a score of 15 on a simple spelling test is not comparable to a score of 15 on a difficult one (e.g., Test A vs Test B in Table 3-1).
Frames of Reference
The meaning derived from scores depends on the frame of reference used for comparison.
Three dimensions define this frame of reference:
Temporal Dimension: Current skills vs. future abilities.
Performance Type: Maximum performance (correct answers) vs. typical performance (how often tasks are completed).
Comparison Standard: Test content, personal past performance, or group average.
Types of Evaluations
Summative Evaluation: Measures accomplishments at the end of a period (e.g., report cards).
Formative Evaluation: Assess skills to guide ffuture instruction.
Achievement Tests
Measure what has been learned.
Types include domain-referenced and criterion-referenced tests.
Domain-referenced tests focus on mastery of specific knowledge.
Norm-Referenced Evaluation
Evaluates performance based on comparison with normative data or groups.
Types of norms include Grade Norms, Age Norms, Percentile Norms, and Standard Score Norms.
Cautions in Using Norms
Norms do not indicate absolute achievement.
An individual can be high relative to a group but still low in absolute terms.
Educational performance should not solely depend on student comparisons; factors such as learning environment should also be considered.
Item Response Theory (IRT)
Focuses on the characteristics of individual items on a test to provide information about a person’s abilities.
Uses curves to indicate probability of a correct response.
Promotes computer adaptive testing (CAT) for personalized assessment.
Summary
Scores are understood through context and comparison, either criterion-based or norm-based.
Diverse scoring methods (e.g., age norms, grade equivalents) allow differentiated analysis for setting educational targets and standards.
Key Terms:
Criterion-Referenced Evaluation: Focuses on performance against a standard.
Norm-Referenced Evaluation: Focuses on performance in relation to others.
Developmental Standard Scores: Interval scores that provide estimates of ability.
Percentile Ranks: Indicates percentage of the norm group that scores below a certain score.
Z-scores: Standard deviations from the mean indicating individual performance.
Examples and Tables referenced include:
Table 3-1: Two spelling tests differences.
Table 3-4: Determining Percentile Ranks.
Table 3-6: Standard Scores and Norms.
Figures showcasing grade equivalent comparisons and profiles.