The Cornell Critical Thinking Test (Level X) is a standardized test designed to assess critical thinking skills in adults and advanced secondary students. It was developed in the 1990s at Cornell University by Robert Ennis and other researchers.
The test aims to measure abilities related to identifying arguments, assumptions, conclusions, alternatives perspectives, making inferences using inductive or deductive reasoning, judging credibility of sources, detecting bias, and other higher-order thinking skills beyond lower-level memorization or comprehension. It does not test specific subject matter knowledge, but rather more general and transferable cognitive processing skills when presented with new information across a variety of topics.
The Cornell test consists of 52 multiple choice items, with each item presenting a short passage of text ranging from a couple sentences to a brief paragraph in length. Test-takers are asked to analyze reasoning, make inferences, weigh evidence, assess credibility, consider other viewpoints, or apply logical principles to each passage scenario. They then select the best answer from 4 to 5 response options for each item. No outside materials or reference sources are permitted during the test.
Administration and scoring of the test is standardized. It is untimed but typically takes 60-90 minutes for most test-takers to complete. Raw scores are then converted to standardized scores with a mean of 50 and standard deviation of 10, allowing comparisons to average critical thinking abilities of adults according to available norms data. Test forms are equated so results can be interpreted equally regardless of which specific version was administered.
Evidence supporting the construct validity and reliability of the Cornell test has been established through numerous research studies examining its psychometric properties since its initial development. Content validity is supported by the test’s alignment to accepted conceptual frameworks of critical thinking as purposeful, reasoned, and goal-directed thinking.
To assess construct validity, numerous factor analytic studies have found the test items to load significantly onto factors reflecting core abilities consistent with theoretical models of critical thinking rather than general intelligence or memory skills alone. Correlations with other measures of analytic reasoning provide further evidence the test measures what it aims to, and not simply broad cognitive ability.
Internal consistency reliability studies have consistently found high alpha coefficients ranging from .80 to .87 for the full test across various samples and administrations. Test-retest reliability correlations have also been shown to be good to very good over short and longer retest intervals. Inter-scorer reliability is not an issue as the test employs multiple-choice scoring.
Concurrent validity evidence for the Cornell test comes from moderate to strong relationships found through correlation with other tests designed to measure similar reasoning constructs such as the Watson-Glaser Critical Thinking Appraisal, California Critical Thinking Skills Test, and performance on certain subscales of intelligence tests involving inductive/deductive reasoning and verbal comprehension.
Discriminant validity studies show it differentiates between groups expected to differ in critical thinking ability based on factors like educational level, grades, or prior academic experiences. For example, college students outperform those with lower education, and gifted students best non-gifted controls. Relationships with grade point average provide further predictive validity support, though the correlations are modest.
Cornell Critical Thinking Test is considered one of the more rigorously developed instruments for assessing critical reasoning and complex cognitive processing skills. Since it is a multiple-choice format, some argue it may fail to capture the full range or depth of critical thinking that could be demonstrated through open-ended or performance tasks.
The test’s abilities are also somewhat narrow in focusing on cognitive processing of brief text or stimulus materials without addressing domain-specific critical thinking or evaluating skills used when problems are ill-defined. These limitations should be taken into account when deciding if it appropriately measures intended constructs or outcomes for a given applied purpose or population.
With reasonable supports for both reliability and construct validity based on extensive research, the Cornell Critical Thinking Test remains one of the most widely established standardized critical thinking assessments used in education research and to evaluate programs aiming to enhance higher-order reasoning abilities in young adults and older adolescents. It should never be the sole measure and must be considered within its limitations to ensure valid interpretation of results.
Andrew Stroud