norms

Norms and Interpretation of Test Scores

Chapter V Overview

  • Importance of norms in psychological testing, particularly in understanding how test scores are interpreted relative to peers.

Norm Definition

General Definition

  • Norms: Unwritten rules or standards of behavior that are deemed acceptable within a particular culture or social group.

Norms in Psychological Testing

  • In the context of psychological testing, norms refer to the scores obtained from standardized tests administered to a representative sample from the target population.

  • These norms provide a baseline for evaluating individual performance in comparison to typical scores expected for a specific grade level or demographic.

Understanding Norms

Performance Expectations

  • Norms express performance expectations and outcomes for defined groups on assessments, comparing results against a distribution of scores from the sample population.

  • Common types of norms used include:

    • Mean: The average score across a sample.

    • 50th Percentile: The midpoint score representing the score below which 50% of the observations fall.

Score Comparisons

  • Norms play a crucial role in evaluating an individual’s performance against a standardization sample, enabling educators and psychologists to assess relative strengths and weaknesses.

Types of Norm-Referenced Tests

  • Norm-Referenced Tests (NRTs) compare an individual’s progress against peers. These tests can be tailored to various categories, such as grade level, age group, and disability status.

  • Norms are derived from the distribution of test scores within normative groups, which are established through rigorous sampling methods.

  • Example: A test of anxiety specifically designed for college students helps establish relevant norms, thus guiding the interpretation of individual scores in the context of typical student experiences.

Score Interpretation

Example of Interpretation

  • When evaluating test scores, an average score for normative groups is often determined (e.g., a score of 19).

  • If an individual's score is significantly higher than this norm (e.g., 24), it indicates a performance level above average in the variable being measured (like test anxiety).

Examples of Norm-Referenced Tests

  • Common examples of standardized tests include the SAT, various IQ tests, and tests graded on a curve, such as the National Assessment Test (NAT) and the National Composite Examination (NCAE).

  • Percentile ranks generated from these tests provide essential comparative data, illustrating where an individual scores relative to others (e.g., scoring in the 80th percentile means performing better than 80% of test-takers).

Norm vs. Criterion Referenced Tests

Norm-Referenced Tests (NRTs)

  • NRTs evaluate an examinee’s performance against a broader group. They are generally designed for ranking individuals and aiding decision-making in educational and professional settings, such as college admissions.

Criterion-Referenced Tests (CRTs)

  • CRTs assess performance against a predefined set of criteria or learning standards. These tests often focus on measuring mastery of specific skills, with results typically leading to a pass/fail outcome.

  • Example: The National Board Medical Exam serves to evaluate whether candidates demonstrate the necessary medical skills adequately, emphasizing mastery over comparison.

Characteristics of Assessments

Norm-Referenced Tests

  • Critics of NRTs argue that these assessments can foster unnecessary competition among peers, potentially skewing performance perceptions and emphasis on standardized comparisons.

Criterion-Referenced Tests

  • CRTs aim for diagnostic and instructional uses, focusing on individual skill development rather than competitive rankings.

Normal Distribution and the Bell Curve

Understanding Distribution

  • Norm-referenced tests are typically aligned with a bell-shaped distribution, which visually represents the clustering of scores around an average, with fewer individuals achieving exceptionally high or low scores.

Statistical Framework

Use of Statistics

  • Descriptive statistics are employed to summarize evaluations relative to a larger population and infer broader insights (e.g., average performance of a class).

  • Inferential statistics enable researchers to draw conclusions about populations based on sample data, thus providing insights into overall trends and patterns.

Scales of Measurement

Measurement Definition

  • Measurement: The application of established rules for assigning numbers to objects or characteristics based on defined criteria.

  • Different scales of measurement arise from the nature of data being assessed:

    • Nominal: Used to categorize items without any inherent order or ranking (e.g., hair color, gender).

    • Ordinal: Ranks individuals or objects based on qualitative attributes but does not allow for scale differences (e.g., satisfaction ratings).

    • Interval: Includes equal intervals between values but lacks an absolute zero (e.g., temperature).

    • Ratio: All properties of interval measures but includes an absolute zero, allowing for a meaningful comparison of magnitudes (e.g., height).

Properties of Scales

  • Key properties inherent in measurement scales include:

    • Magnitude: Indicates the comparison of more or less distinctions between ranks.

    • Equal Intervals: Ensures consistent meanings across differences in the scale.

    • Absolute Zero: Represents the complete absence of the characteristic being measured.

Examples of Scales

  • Nominal Scale Examples: Hair color, gender, and other categories.

  • Ordinal Scale Examples: Satisfaction ratings that scale from strong agreement to strong disagreement.

  • Interval and Ratio Scale Examples: Temperature measurements (interval) and height measurements (ratio).

Standard Scores and Their Significance

Definition

  • Standardized scores, such as Z scores, help in making comparisons across a normed dataset.

  • These scores indicate how many standard deviations a particular score is from the mean, providing a relative perspective.

Understanding Distribution of Scores

  • The properties of normal distribution are utilized to interpret how a score ranks compared to others in the sample.

  • Z-scores inherently illustrate the performance of individuals relative to their peer group in terms of standard deviations.

Advanced Score Types

  • T Scores: Commonly used when sample sizes are small and the population standard deviation is unknown, providing a standardized metric for assessing performance.

  • Stanine Scores: A scale from 1 to 9 used for representing relative performance; scores of 7-9 are considered good, 4-6 average, and 1-3 below average.

Percentages, Percentiles, and Stanines

Definitions

  • Percentage: Represents a part per hundred of a certain attribute.

  • Percentile: Denotes the rank positioning in comparison to others below a specific score, highlighting relative standing among scores.

  • Stanine: A standardized test scoring method that categorizes scores on a nine-point scale, with a meaningful average centered around the mean.