THE CENTRAL IDEA
A empirical evaluation programme must distinguish what people say they do from how they reason in defined tasks.
Concept and evidence
MDP’s evidence programme should begin with the narrow claim it makes: that its items describe distinguishable reported practices. Content review and cognitive interviews would test whether respondents understand the intended actions. Later studies could compare responses with observed behavior in decision tasks. Research on competence and forecasting provides examples of performance measurement, but MDP must not borrow those instruments’ measurement evidence or imply equivalence to their scores.
The criteria should match the proposed domain. Evidence appraisal might be compared with whether a participant checks a consequential claim, while uncertainty communication might be compared with how they describe the limits of an estimate. These observations still do not establish broad decision quality. Outcome measures would need their own rationale, and researchers should examine whether verbal fluency, prior knowledge, or social desirability explains the association.
Applied interpretation and an original worked example
An illustrative study could present volunteers with a low-stakes planning problem containing incomplete information. Participants would explain their approach, and independent raters would code specific actions without seeing MDP results. The study would preregister which relationships are expected and distinguish exploratory findings. A later sample would test the revised model rather than reuse the same data to claim confirmation.
A separate feedback study could examine whether an MDP report helps people produce a clearer decision record or choose a proportionate review process. It should also test unwanted effects, such as unnecessary analysis, overconfidence, or the belief that a favorable profile proves immunity to bias. Satisfaction is only one outcome. The research programme should remain open to finding that some dimensions overlap or that a shorter framework is more useful. Scientific credibility depends on reporting those results honestly, not on preserving six dimensions because the report design looks balanced or because the names are attractive to prospective buyers.
Designing criteria that do not duplicate the questionnaire
A empirical evaluation study would be circular if it asked people whether they use careful reasoning and then judged measurement evidence from another questionnaire asking nearly the same thing. MDP needs criteria that illuminate the proposed practices from a meaningfully different angle. A low-stakes task could preserve which information a participant requested, which alternatives they considered, and what they wrote before learning the outcome. Those observations remain incomplete, but they are different from another global self-description. Reviewers would need clear coding rules and should not know the participant's inventory result while applying them.
The study should also separate reasoning process from reading fluency, task familiarity, and subject knowledge. A participant may identify a crucial issue quickly because they have seen the task before; another may use a careful routine yet lack the knowledge needed to solve it. Researchers could examine performance across more than one domain and specify in advance what kind of relationship would support each dimension's interpretation. They should include competing explanations and an independent replication after any item revision. If only one general factor survives, or if narrative fluency explains most of the associations, six separate decision-practice scores would require reconsideration. The aim is to make the proposed claims vulnerable to evidence rather than to design a demonstration that must succeed.
Points to carry forward
- Use independent, task-specific criteria and test alternative explanations.
Where the evidence stops
No MDP behavioral empirical evaluation, population reference distributions, or predictive study has been completed.
No established reliability, population reference distributions, prediction evidence, or professional endorsement is claimed.
The cited literature informs our original framework. Read the current evidence status and intended use alongside this guide.
REFERENCES / FOLLOW THE ORIGINAL EVIDENCE