EPI flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Hoshin Kanri

A strategy deployment method that translates top-level breakthrough objectives into successively more specific divisional, departmental, and daily management targets, each with a named owner, metric, and review cadence.

Project charter minimum content

A charter states the problem, quantified baseline, target, scope boundaries, sponsor, team, milestones, and resource commitment. Without a quantified baseline the project cannot demonstrate improvement or be approved.

Benefit realization period

The post-handoff window, typically six to twelve months, during which the operational owner confirms that gains have held before a project is formally closed and savings are booked.

Decision rights matrix

A RACI or DACI table assigning who recommends, approves, is consulted, and is informed for each decision. Its main purpose is ensuring every metric has exactly one accountable owner.

Operational definition

A written specification of numerator, denominator, data source, inclusion and exclusion rules, frequency, and owner for a measure, so that independent observers collect the same value.

Run chart shift rule

Six or more consecutive points on the same side of the median constitutes a non-random shift, indicating a process change rather than ordinary common cause variation.

Control limits versus specification limits

Control limits are calculated from observed process data, typically the mean plus or minus three sigma. Specification limits come from customer or design requirements and belong on capability studies, not control charts.

Common versus special cause

Common cause variation is inherent in the process and calls for redesign. Special cause variation comes from a specific identifiable event and calls for investigation. Treating common cause as special is tampering.

Balancing measures

Metrics that detect unintended consequences of an improvement, such as rising cost, reduced access, or increased staff workload, and are monitored alongside the primary outcome measure.

SIPOC diagram

A high-level map listing Suppliers, Inputs, Process, Outputs, and Customers. It fixes project scope and identifies the customer before detailed process mapping begins.

Flow efficiency

Value-added processing time divided by total lead time. Transactional and administrative processes commonly fall below ten percent, meaning most elapsed time is waiting.

Little's Law

Average work in process equals arrival rate multiplied by average cycle time. It predicts how queues and cycle time respond to changes in workload or capacity.

Utilization and waiting time

Waiting time rises nonlinearly as utilization approaches one hundred percent. Systems above roughly eighty to eighty-five percent utilization become fragile, and small variation produces long delays.

Bottleneck (constraint)

The process step with the highest utilization or longest queue. System throughput is limited by this step, so improving any non-constraint does not increase overall output.

PDSA cycle

Plan, Do, Study, Act: a rapid small-scale test of a change, used to learn quickly before full implementation. It suits iterative improvement rather than definitive root cause analysis.

DMAIC phases

Define, Measure, Analyze, Improve, Control. Define sets the charter, Measure establishes the baseline, Analyze finds root causes, Improve pilots solutions, and Control sustains the gain.

Poka-yoke

An error-proofing design that makes a mistake impossible or immediately visible, such as a connector that fits only one way or a forced-field form entry.

Process capability Cpk

A ratio comparing the distance from the process mean to the nearest specification limit against three sigma. Values below one indicate output falling outside specification.

FMEA risk priority number

Severity multiplied by occurrence multiplied by detection, each scored one to ten. High severity items are prioritized regardless of the total, because detection controls cannot reduce severity.

Measurement system analysis

A study of repeatability and reproducibility that quantifies how much observed variation comes from the measurement system. It is performed before capability analysis or improvement conclusions.

Kotter's change sequence

Eight steps from establishing urgency through forming a coalition, creating vision, communicating, removing barriers, generating short-term wins, consolidating gains, and anchoring changes in culture.

ADKAR model

An individual change model requiring Awareness, Desire, Knowledge, Ability, and Reinforcement. Change fails when any element is missing, and Desire is the most frequent constraint.

Lewin unfreeze-change-refreeze

A change model in which existing norms are loosened, the change is implemented, and new behaviors are locked in through standard work, training, metrics, and job descriptions.

Resistance as diagnostic signal

Opposition to change usually indicates missing information, missing capability, threatened status, or legitimate technical disagreement. Each cause requires a different response rather than generic persuasion.

Analytics maturity ladder

Four levels: descriptive reporting of what happened, diagnostic explanation of why, predictive forecasting of what will happen, and prescriptive recommendation of what action to take.

Interrupted time series design

An evaluation design that models the pre-intervention trend and tests for change in level and slope after the intervention. It is the standard quasi-experimental method for improvement evaluation.

Statistical test selection

Continuous outcomes comparing two groups use a t-test, three or more groups use ANOVA, and categorical associations use chi-square. Non-normal distributions call for rank-based alternatives such as Mann-Whitney.

Model calibration versus discrimination

Discrimination measures how well a model ranks cases, often by area under the curve. Calibration measures whether predicted probabilities match observed rates. Decision use requires both.

Stakeholder influence-interest grid

A two-by-two plot of influence against interest that assigns engagement strategies: manage closely, keep satisfied, keep informed, or monitor, depending on each stakeholder's position.

Closing the feedback loop

Reporting back to participants what was heard and what changed as a result. Failing to close the loop reduces willingness to participate in future engagement efforts.

SBAR communication

Situation, Background, Assessment, Recommendation: a structured briefing format that leads with the decision needed and the specific ask rather than with narrative history.

Minimum necessary standard

A privacy principle limiting collection, use, and disclosure of personal data to what is required for the stated purpose, supported by role-based access, audit logs, and retention schedules.