Module 2.8b Intelligence and Achievement: Assessing Intelligence

Intelligence Test: A method for assessing an individual's mental aptitudes and comparing them with those of others, using numerical scores.
Achievement Test: A test designed to assess what a person has learned.
Aptitude Test: A test designed to predict a person's future performance; aptitude is defined as the capacity to learn.

A correlation coefficient of +.82 indicates a strong positive relationship between SAT scores and intelligence test scores, suggesting that as the scores on one variable increase, scores on the other variable increase as well.
The correlation between ACT scores and intelligence is +.77 (Koenig et al., 2008).
Figure 2.8-3: Displays a scatterplot showing the close correlation between intelligence test scores and verbal and quantitative SAT scores (Data from Frey & Detterman, 2004).

Explain the characteristics of an intelligence test, and compare and contrast achievement tests and aptitude tests.

Explain when and why intelligence tests were created, and how today's tests differ from early intelligence tests.

Explain the elements of the normal curve, and explain standardization, reliability, and validity.

Intelligence tests assess mental aptitudes and compare them with others using numerical scores.

Achievement Tests: Reflect what has been learned (e.g., AP exams).
Aptitude Tests: Predict potential learning abilities (e.g., college entrance exams).

Francis Galton: Laid the groundwork for assessing intelligence by attempting to measure human traits through reaction time, sensory acuity, and muscular power, leading to the eugenics movement.
Alfred Binet: Developed the first modern intelligence tests in France to evaluate children's learning potential, aiming to minimize bias and improve educational opportunities.

Mental Age: A measure developed by Binet to indicate the level of performance associated with children of a certain chronological age. For example, a child performing at the level of an average 8-year-old is given a mental age of 8.
Stanford-Binet Test: Lewis Terman's adaptation of Binet's tests in America, establishing an intelligence quotient (IQ) calculated as:
$IQ = (\text{Mental Age} / \text{Chronological Age}) \times 100$

David Wechsler: Created the Wechsler Adult Intelligence Scale (WAIS) and its child versions to assess intelligence more broadly, incorporating various subtests to measure different cognitive abilities, such as:
- Similarities: Commonality reasoning between objects/concepts.
- Vocabulary: Identifying and defining words.
- Block Design: Visual abstract processing.
- Letter-Number Sequencing: For auditory processing and memory.

Standardization: Establishes uniform testing procedures by administering the test to a representative sample; comparison with this sample allows for meaningful interpretation of scores.
Reliability: A reliable test yields consistent results across multiple administrations (split-half methods, alternate forms, retesting). Examples include the Stanford-Binet, WAIS, and WISC tests showing high reliability (correlation coefficients of about +.9).
Validity: A test's ability to measure or predict what it claims to.
- Content Validity: Measures pertinent behavior (e.g., driving tests).
- Construct Validity: Measures a specific trait or concept (self-esteem measurement).
- Predictive Validity: Indicates how well test scores predict future performance (e.g., SAT and GRE correlations).

Normal Curve: A bell-shaped graph illustrating score distribution in intelligence testing, where the average score is designated as 100.
Approximately 68% of the population scores within one standard deviation (about 85 to 115).
The Flynn Effect: Refers to the observed rise in intelligence test performance over time across cultures, with average IQ scores increasing about three points per decade due to factors such as improved nutrition, education, and access to information.

To evaluate a test's reliability, researchers may use:
- Test-Retest: See if scores remain consistent over repeated tests.
- Split-Half: Evaluate consistency between different halves of the same test.
- Alternate Forms: Compare results of different forms of the test.
- Tests must have a high correlation among scores to be deemed reliable.

Content Validity: Ensures a test samples the appropriate behavior.
Predictive Validity: Assesses how well a test predicts future performance by correlating scores with actual performance over time.

Predictive power diminishes as students progress through schooling due to restricted score ranges, which can reduce the correlation between aptitude tests and academic performance, as seen with inflation in grades over time.

Intelligence tests assess and compare mental abilities using standardized numerical scores, with distinctions made between achievement and aptitude tests.
The evolution of intelligence tests shows a transition from early eugenics-focused approaches to modern assessments that acknowledge environmental influences and seek to improve educational outcomes without restrictive labeling.
Tests must meet standards for standardization, reliability, and validity to be considered useful and informative in assessing cognitive abilities.