Your defaults. Portable.
Communication preferences are adjustable defaults, not a diagnosis or a learning-style classification.
Dialed asks which responses you would prefer, then turns those choices into editable instructions. The questions and mappings are original product choices. The assessment has not been scientifically validated and does not measure ability, personality, or a fixed learning style.
Brevity, warmth, diagrams, and analogies are communication choices. Finding an answer pleasant or useful does not establish better learning, retention, or coding performance. Accessibility accommodations are explicit choices, independent of your answers.
These papers help frame the design. Their findings concern the people and tasks studied; none tested Dialed or its exported instructions.
Pashler and colleagues reviewed learning-style claims across educational settings. They found insufficient evidence to justify assigning styles and matching instruction to them. This does not mean every kind of personalization is ineffective.
Dialed therefore offers adjustable defaults without assigning learner types or claiming that a preferred format improves learning.
Pashler et al. — Learning Styles: Concepts and Evidence (2008 issue)Dunlosky and colleagues’ 2013 review compared ten techniques across ages, prior knowledge, materials, and learning outcomes. Practice testing and spreading study over time received high utility ratings; self-explanation received a moderate rating with evidence gaps.
This informs our interest in optional practice and explaining reasoning. It does not prove that one AI question, explanation, or preference profile produces those benefits. Dialed schedules no practice or reminders.
Dunlosky et al. — Improving Students’ Learning With Effective Learning Techniques (2013)In two experiments with university students learning short prose passages, Roediger and Karpicke found that recall tests without feedback outperformed repeated study on delayed retention tests. Repeated study did better at five minutes.
Optional recall practice is a design possibility, not evidence that Dialed improves programming skill or that every request should become a quiz. Interactive practice requires your explicit request.
Roediger & Karpicke — Test-Enhanced Learning (2006)Sweller’s 1988 paper combined a computational model with high-school mathematics experiments to examine how goal-directed problem solving can consume capacity needed to learn underlying patterns.
Worked explanations and manageable sections are design hypotheses inspired by this account. It does not establish an ideal response length, diagnose cognitive load, or justify forcing small steps on everyone.
Sweller — Cognitive Load During Problem Solving (1988)Sources reviewed 2 October 2026. These are bounded summaries of foundational papers, not a current systematic review of AI tutoring.
There are 12 core questions and three optional calibration questions. They use everyday work and learning situations, including coding examples; no programming knowledge is needed. You can skip any item.
Results show “Why this?” answer evidence and let you edit or reset each preference. Your edits survive re-answering until you reset them. Presets are shortcuts to settings; imported profiles cannot reconstruct answer explanations. Neither invents assessment answers.
You have one set of editable defaults with work and learning rules. Your current request overrides stored preferences; small steps never require a pause on their own. The comparisons are authored examples, and the exports are compiled locally without runtime AI generation.
Review and edit your defaultsWe have not established the questionnaire’s reliability, an improvement in learning outcomes, or a reduction in repeated prompting. Usability and perceived fit need their own evaluation; learning efficacy would require a separate comparison measuring correctness, transfer, and retention. Enjoyment and installation success cannot substitute for those outcomes.
A host may not activate or follow every instruction. See the installation guide and compatibility limits before relying on an export.