Connectionist Models in Cognition

Symbolic cognitive models operate on the principle that cognition is computation involving the manipulation of symbols. These models utilize high-level, explicit representations like propositions, schemas, and rules (e.g., IF-THEN statements) to simulate structured thought processes such as logical reasoning, language use, and problem-solving. They form the basis of classical artificial intelligence, often called 'Good Old-Fashioned AI' (GOFAI).

The System Usability Scale (SUS) is a widely used, reliable 10-item questionnaire for assessing perceived usability. Users rate each statement on a 5-point Likert scale. The responses are then converted into a single score from 0 to 100. Despite its simplicity, it provides a quick and robust measure of a system's overall usability from the user's subjective viewpoint.

The Kruithof curve is an empirical graph from psychophysics that outlines a region of illuminance levels and color temperatures considered comfortable or pleasing. It postulates that at low light levels, humans prefer warmer color temperatures (more reddish/yellowish), while at high illuminance levels, cooler color temperatures (more bluish) are preferred. Lighting conditions outside this region may feel unnatural or unpleasant.

A design philosophy and iterative process where the needs, wants, and limitations of end-users are given extensive attention at each stage of the design process. UCD aims to create products that are highly usable and accessible by involving users throughout the design cycle through research, testing, and feedback, rather than relying solely on designer assumptions.

ACT-R (Adaptive Control of Thought—Rational) is a cognitive architecture defining the basic, fixed structures of the human mind. It models cognition as a set of independent modules, such as perceptual, motor, and declarative memory, that interact through a central production system. Information is held in buffers, and production rules fire to manipulate this information, simulating human problem-solving and learning processes.

Bayesian models of cognition frame the mind as a probabilistic inference engine. This approach posits that the brain represents knowledge as probability distributions and updates these beliefs upon receiving new evidence according to Bayes' theorem. It models perception, learning, and reasoning as optimal or near-optimal statistical inference under uncertainty, providing a unifying mathematical framework for many cognitive functions.