Psychological Testing Overview

Psychological Testing Definition

• Psychological testing refers to the use and application of psychological and educational tests.
• The main purpose of psychological testing is to evaluate individual differences between people.
• A psychological test is a measurement tool that quantifies behavior and helps understand or predict behavior.
• A test item is a specific stimulus to which a person responds overtly.

Types of Tests

• Individual tests are administered to one person at a time.
• Group tests are administered to multiple people at the same time by a single examiner.
• Ability tests measure skills in terms of speed, accuracy, or both.
  • Achievement tests measure previous learning.
  • Aptitude tests measure potential for learning a specific skill.
  • Intelligence tests measure potential to solve problems, adapt to change, and learn from experience.
• Personality tests measure typical behavior, including traits, temperaments, and dispositions.
  • Structured (objective) personality tests use self-report statements requiring a "True" or "False" or "Yes" or "No" response.
  • Projective personality tests provide ambiguous stimuli with unclear response requirements.

History of Psychological Testing

• Evidence suggests the Chinese used a civil service testing program over 4,000 years ago.
• The Han Dynasty commonly used test batteries for candidate selection for the state bureaucracy.
• Western psychological testing usage can be traced back to the 19th century.
• Charles Darwin's theory of evolution and concept of individual differences had a significant impact on the development of psychological testing.
• Sir Francis Galton, Darwin's cousin, applied Darwin's theories to the study of humans and started research into human individual differences.
• German psychophysicists, including Herbart, Weber, Fechner, and Wundt, used mathematical models to study the mind and sensation.
• Alfred Binet developed the first intelligence test, the Binet-Simon Scale, to identify intellectually subnormal individuals.
• The Binet-Simon Scale introduced the concept of mental age and underwent multiple revisions, including the Stanford-Binet Intelligence Scale.
• World War I led to the development of group intelligence tests, such as the Army Alpha and Army Beta, to assess recruits.
• The period between World Wars I and II saw the development of personality tests to measure enduring characteristics or traits.
• Structured personality tests like the Woodworth Personal Data Sheet emerged, followed by projective tests like the Rorschach inkblot test and the Thematic Apperception Test (TAT).
• The Minnesota Multiphasic Personality Inventory (MMPI) marked a shift towards using empirical methods to determine the meaning of test responses.
• Factor analysis, a statistical technique to identify underlying dimensions or factors, was used to develop tests like the Sixteen Personality Factor Questionnaire (16PF).

Types of Scales

• Scales of measurement are used to assign numbers to objects that represent the magnitude of the attribute being measured.
• Magnitude represents the property of "moreness."
• Equal intervals mean the difference between two points on the scale has the same meaning.
• An absolute 0 is achieved when there is nothing of the property being measured.

Statistical Concepts Related To Testing

• Descriptive statistics provide a concise description of a collection of numerical information.
  • Frequency distributions display scores to show how often each value was obtained.
  • Percentile ranks show the percentage of scores that fall below a specific score.
  • Mean (average) is the sum of a set of scores divided by the number of scores.
  • Standard deviation measures the approximation of the average deviation from the mean.
  • Variance measures the average squared deviation around the mean.
• Inferential statistics are used to draw inferences from observations of a sample to a larger population.
• Z-score transforms data into standardized units by indicating the distance of a score from the mean in standard deviation units.

Methods of Correlation

• Correlation is a statistical technique that measures the direction and magnitude of the relationship between two variables.
• A correlation coefficient is a mathematical index that describes the direction and magnitude of a relationship.
• The Pearson product moment correlation coefficient is a ratio used to determine the degree of variation in one variable that can be estimated from knowledge about variation in the other variable.

Reliability

• Reliability refers to the accuracy, dependability, consistency, or repeatability of test results.
• Reliability refers to the degree of test scores being free of measurement errors.
• Measurement error is the discrepancy between an individual's true score and their observed score due to factors that can affect test performance.
• Test–Retest method assesses the consistency of test results when the test is administered at different times.
• Parallel forms method compares two equivalent forms of a test measuring the same attribute.
• Internal consistency methods examine how people perform on similar subsets of items selected from the same form of the measure.
  • Split-half method divides the test into halves and compares performance on the two halves.
  • KR20 formula calculates the reliability of a test with dichotomous items (scored 0 or 1).
  • Coefficient alpha is a general reliability estimate for non-dichotomous items.

Validity

• Validity refers to the meaning and usefulness of test results.
• Validity is the degree to which a certain inference or interpretation based on a test is appropriate.
• Face validity is the mere appearance that a measure has validity, not a real type of validity evidence.
• Content-related evidence considers the adequacy of representation of the conceptual domain the test is designed to cover.
• Criterion-related evidence shows how well a test corresponds with a specific criterion.
  • Predictive validity is the test's ability to forecast future performance on a criterion.
  • Concurrent validity is the test's relationship with a criterion measured at the same time.
• Construct-related evidence is established by demonstrating the relationship between a test and other tests and measures.
  • Convergent evidence shows when a measure correlates well with other tests believed to measure the same construct.
  • Discriminant evidence is when a test has low correlations with measures of unrelated constructs.

Guidelines for Test Writing

• Clearly define what you want to measure.
• Use substantive theory as a guide.
• Make items as specific as possible.
• Generate an item pool.
• Avoid redundant items.
• Avoid exceptionally long items.
• Keep the reading difficulty level appropriate for the test takers.
• Avoid "double-barreled" items that convey two or more ideas at the same time.
• Consider using the item characteristic curve to evaluate the performance of items graphically.

Test Formats

• Dichotomous format (True-False):
  • Advantages: Simplicity, ease of administration, quick scoring, requires absolute judgment.
  • Disadvantages: Encourages memorization, susceptible to guessing.
• Polytomous format (Multiple-Choice):
  • Advantages: Easy to score, lower probability of guessing correctly, quick to answer, covers a large amount of information.
  • Disadvantages: Difficult to find good distractors, may not increase reliability.
• Essay format:
  • Advantages: Can assess higher-order thinking skills, requires in-depth understanding.
  • Disadvantages: Difficult to score objectively, time-consuming to grade, may not be reliable if scoring procedures are not standardized.
• Likert format:
  • Advantages: Easy to use, familiar to respondents.
  • Disadvantages: Susceptible to response biases.
• Category format (Rating Scales):
  • Advantages: Provides a wider range of responses, can measure subtle differences.
  • Disadvantages: Susceptible to response biases, difficult to interpret if intervals are not equal.

Stereotype Threat

• Stereotype threat happens when individuals from stereotyped groups are concerned about confirming negative stereotypes about their group, which can lead to underperformance on tests.
• Stereotype threat can affect performance by:
  • Triggering anxiety and negative thoughts, leading to distraction and reduced attentional capacity.
  • Depleting working memory, as individuals try to suppress interfering thoughts related to the stereotype.
  • Encouraging self-handicapping, where individuals reduce their effort to provide an excuse for poor performance.
  • Causing physiological arousal, which can interfere with performance on challenging tasks.

Individual and Group Tests

• Individual tests are administered to one person at a time by a trained examiner allowing for observation and interaction.
• Individual tests are used for clinical assessment, such as intelligence testing and neuropsychological evaluation.
• Group tests can be administered to multiple people simultaneously.
• Group tests are typically more efficient and cost-effective for large-scale assessments like educational testing and personnel selection.

Strategies Used For Test Construction

• Deductive strategies use reason and deductive logic to develop personality measures.
  • Logical-Content strategy: Items are derived based on reason and deductive logic.
  • Theoretical strategy: Items are derived based on a specific theory about the nature of the characteristic being measured.
• Empirical strategies rely on data collection and statistical analyses to determine the meaning of test responses.
  • Criterion-Group strategy: Items are selected to discriminate between a criterion group (individuals who share a characteristic) and a control group.
  • Factor Analytic strategy: Uses factor analysis to determine the underlying dimensions or factors measured by a set of items.