EDAC common mistakes
EDAC · Understand the traps and how to avoid them.
When a business-case item asks whether an EBD investment 'pays off,' the candidate computes only first/construction cost and answers with a capital budget figure, never converting the targeted outcome (e.g., reduced falls, shorter length of stay, lower staff turnover) into an annual dollar value, a payback period, or a lifecycle comparison.
Most candidates are trained in project budgeting, where 'cost' means construction cost; lifecycle cost, simple payback, ROI and net present value are finance vocabulary they have rarely had to use in a design conversation, so under time pressure they default to the number they know.
What to do instead
Drill a four-line business-case template until it is automatic: (1) design intervention and its incremental first cost, (2) the operational outcome it is meant to change, (3) that outcome monetized per year using the client's own unit costs, (4) simple payback = incremental cost / annual savings. Practise doing this for two or three classic outcome categories before exam day.
Treats evidence-based design as a citation-gathering step performed after the design concept is already fixed - i.e., the candidate describes 'finding studies that support our design' rather than a cycle that begins with explicitly stated goals and ends with measurement and sharing of results. Given a scenario, they cannot say which step of the process the project is actually in.
In everyday practice, research is often retrofitted to justify a decision, and firms rarely complete the loop by measuring post-occupancy outcomes; the post-rationalization habit feels like EBD and is easy to mistake for it.
What to do instead
Memorize the process as an ordered loop - define goals and objectives, find and critically appraise evidence, translate into design concepts, implement, measure outcomes, disseminate - and for every practice scenario force yourself to name the current step and the next one. If a scenario has a design fix but no stated goal or measure, that is the tell that the process was skipped.
Writes or accepts objectives such as 'create a calming, patient-centred environment' or 'improve staff satisfaction' as if they were measurable, and cannot say what indicator, direction, or time frame would show success.
Design briefs and vision statements are written in aspirational language, so these phrases sound normal; candidates reproduce the client's language instead of converting it into something a study could test.
What to do instead
Run every objective through a metric-direction-timeframe filter: 'Reduce (direction) patient-reported noise disturbance on the unit (metric) by 20% relative to baseline (timeframe/comparison).' If you cannot name the instrument, unit, or data source that produces the number, it is still a goal, not an objective.
Plans to evaluate outcomes only after occupancy and never establishes a pre-design or pre-occupancy baseline, and never pairs each stated goal with the specific measure and data-collection method that will be used later.
Measurement feels like an end-of-project activity, and baseline data collection competes with design fees and schedule; candidates assume 'before' data will be reconstructible later, which it usually is not.
What to do instead
For each goal, complete a three-column table at the goals stage: goal → measure/indicator → data source and collection timing, and confirm baseline collection is scheduled before construction. Remember the exam logic: an outcome with no baseline is an anecdote.
Misplaces study designs on the evidence hierarchy - ranking a single project case study or an expert opinion above a systematic review or multi-site study, discarding qualitative evidence as 'not real evidence,' or citing a manufacturer white paper or a design guideline as if it were primary peer-reviewed research.
Candidates conflate 'published,' 'award-winning,' and 'peer-reviewed,' and many come from practice cultures where precedent projects and design guides carry more weight than the research literature.
What to do instead
Learn a working hierarchy for design questions (systematic review/meta-analysis → randomized or controlled studies → cohort/case-control → cross-sectional surveys → case studies and post-occupancy reports → expert opinion) and be able to say what each design can and cannot support. Classify every source by study design and by whether it is peer-reviewed primary evidence before using it.
Describes the literature search as keyword searching on a general web engine with no defined question, no structured search string or database selection, and no documented inclusion/exclusion criteria, then cannot explain how the search's rigor would be judged or reproduced.
Candidates have found articles successfully by browsing, so the informal method feels sufficient; they have never had to make a search auditable or defend its boundaries.
What to do instead
Practise framing a design question in a structured format (population, intervention/design variable, comparison, outcome), building a Boolean string with synonyms and controlled vocabulary, choosing appropriate databases and grey-literature sources, and writing explicit inclusion/exclusion criteria with a screening count. Be ready to state that reproducibility and transparency, not hit count, are what make a search credible.
Translates a research finding literally into a numeric design prescription - e.g., 'a study found daylight associated with shorter stays, therefore specify X footcandles' - without checking whether the study population, setting, dose/exposure level, and moderating variables match the project, and without attaching a hypothesis or evaluation plan to the design move.
Clients and design teams want a defensible specification, and a study's headline finding reads like a rule; candidates skip the inferential step between 'association in that context' and 'expected effect here.'
What to do instead
Express every translation as an explicit hypothesis: 'If we do [specific design variable at this level] for [this population in this setting], we expect [this outcome], measured by [this method].' Then state the boundary conditions from the source study (sample, setting, exposure magnitude, confounders) and what would make you wrong.
Uses 'reliable' and 'valid' interchangeably, confuses internal with external validity, or treats a statistically significant result as automatically meaningful for the project and a non-significant one as proof of no effect.
The terms sound like synonyms in ordinary speech, and significance-language is widely misused in popular and professional summaries that candidates have read.
What to do instead
Hold two sentences: reliability = consistency/repeatability of the measurement; validity = whether the instrument or study measures what it claims (with internal validity about causal inference and external validity about generalizability). Separately, always ask whether the observed effect size matters operationally and whether the study was powered to detect it.
Mislabels the purpose and timing of a post-occupancy evaluation - describing it as something done before move-in, as equivalent to a one-time occupant satisfaction survey, or conflating formative evaluation (during design/construction, to adjust) with summative evaluation (after occupancy, to judge outcomes), and confusing a POE with benchmarking against other facilities.
In practice 'POE' is used loosely for any occupant feedback exercise, and the formative/summative distinction is rarely spoken aloud in project meetings.
What to do instead
Anchor POE to three things: it happens after occupancy, it compares measured performance against the project's stated goals/baseline, and it uses multiple methods (observation, environmental measurement, survey, interviews, records data) - not a single questionnaire. Keep a one-line definition of formative vs summative vs benchmarking and match the tool to the question being asked.
Describes stakeholder engagement as a single design charrette or briefing session attended by senior leadership only, with no plan for ongoing involvement of the people who will occupy and operate the building, and no plan for sharing findings with them or with the wider field once the project is measured.
Charrettes are the visible, familiar engagement artefact and they photograph well; operations staff and post-occupancy feedback loops are invisible in a project timeline, so they get omitted.
What to do instead
Map stakeholders to project phases and state who is involved when and for what decision - including facility managers, frontline staff, and (where appropriate) patients/families - and schedule engagement checkpoints rather than one event. Treat dissemination (to the client, occupants, and the professional community) as a required final step of the process, not an optional extra.