Comparison of DiscoTests with conventional tests

DiscoTests and conventional standardized tests can be thought of as complementary. They are designed to test different kinds of skills, and research confirms that they are successful in doing so. Correlations between scores on the kind of developmental assessments made by DTS and scores on conventional multiple choice assessments is in the .40-.60 range. That means that somewhere between 16% to 36% of the kind of learning that is captured by conventional assessments is likely to overlap with the kind of learning that is captured by DiscoTests.

The table below provides a comparison of DiscoTests with conventional standardized tests on a number of dimensions.

Conventional tests
Theoretical foundation Cognitive developmental theory, Dynamic Skill Theory, Test theory Test theory
Scale Fischer’s Dynamic Skill Scale, an exhaustively researched general developmental scale, which is a member of a family of similar scales that were developed during the 20th century. Statistically generated scales, different for each test (though some tests are statistically linked)
Learning sequences Empirical, fine-grained & precise, calibrated to the dynamic skill scale Empirical, coarse-grained and general
Primary item type Open response More or less sophisticated forms of multiple choice
Targeted skills Reasoning with knowledge, knowledge application, making connections between new and existing knowledge, writing Content knowledge, procedural knowledge
Content Carefully selected “big ideas” and the concepts and skills associated with them. The full range of content specified in state standards for a given subject
Educative/formative Yes, (1) each DiscoTest focuses on ideas and skills central K-12 curricula, (2) test questions require students to thoughtfully apply new knowledge and connect it with their existing knowledge, (3) students receive reports with targeted feedback and learning suggestions, (4) teachers learn how student knowledge develops in general and on each targeted concept or skill. Not really, though increasingly claim to be
Embeddable in curricula Yes, DiscoTests are designed to be part of the curriculum. No
Standardized Yes, statistically, calibrated to the skill scale Yes, statistically only
Stakes Low. Selection decisions are based on performance patterns over time on many individual assessments. High. Selection decisions are often based on single assessments.
Ecological validity Direct tests that focus on deepening and connecting knowledge about key concepts and ideas, while developing broad skills that are required in adult life, such as those required for reasoning, communicating, and problem-solving. Tests of proxies, focus on ability to detect correct answers.
Statistical reliability .91+ for a single age cohort (distinguishes 5-6 distinct levels of performance). For high stakes tests, usually .95+ for a single age cohort (distinguishes 6-7 distinct levels of performance).

Leave a Reply

Your email address will not be published. Required fields are marked *