Description

This curriculum spans the rigor and coordination of a multi-workshop quality initiative, covering data practices from initial scoping to long-term governance, comparable to the phased execution seen in enterprise-wide process improvement programs.

Module 1: Defining Data Requirements in the Define Phase

Selecting critical-to-quality (CTQ) metrics based on stakeholder input and project charters
Mapping process outputs to measurable variables that align with business objectives
Establishing operational definitions for each data point to ensure consistency across teams
Identifying data owners and custodians during process scoping to secure access early
Determining data granularity (e.g., per transaction, per shift) based on process cycle time
Deciding between primary and secondary data sources considering accuracy and availability
Documenting data collection constraints such as system access, privacy regulations, or legacy systems
Developing a preliminary data collection plan with timelines and responsibilities

Module 2: Designing Data Collection Tools and Methods

Choosing between check sheets, automated logging, and digital forms based on error rates and volume
Designing paper and electronic forms with built-in validation rules to minimize entry errors
Implementing stratified sampling strategies when population segments behave differently
Setting sampling frequency based on process stability and measurement system capability
Integrating timestamp and operator ID fields to support root cause analysis later
Conducting pilot tests of data collection tools to identify usability gaps
Standardizing units of measure across departments to prevent aggregation errors
Documenting field-level instructions directly on forms to reduce interpretation variance

Module 3: Ensuring Data Accuracy Through Measurement System Analysis

Conducting Gage R&R studies for continuous data with at least 10 parts, 3 operators, 3 trials
Performing attribute agreement analysis for categorical data using kappa statistics
Identifying sources of measurement variation: equipment, appraiser, environment, or procedure
Deciding whether to recalibrate instruments or revise operational definitions based on MSA results
Handling destructive testing scenarios by using split specimens or proxy measurements
Establishing calibration schedules for measurement devices based on usage and drift history
Training data collectors on consistent technique, especially for subjective assessments
Documenting MSA outcomes and obtaining sign-off before proceeding to full data collection

Module 4: Executing Data Collection in the Measure Phase

Deploying trained data collectors with documented protocols and escalation paths
Monitoring real-time data submission rates to detect collection bottlenecks
Implementing data validation checks at point of entry (e.g., range limits, mandatory fields)
Handling missing data: deciding between imputation, exclusion, or re-collection
Logging deviations from the collection plan and justifying adjustments in the project file
Synchronizing data collection across multiple shifts or locations to ensure representativeness
Using barcode scanners or IoT sensors when manual entry introduces unacceptable error
Securing interim data backups and access controls, especially for sensitive process data

Module 5: Validating and Cleaning Data for Analysis

Identifying outliers using statistical methods (e.g., IQR, Z-score) and verifying with process experts
Resolving duplicate records caused by system integration or manual re-entry
Standardizing categorical responses (e.g., “Yes,” “yes,” “Y”) into consistent coding
Handling inconsistent timestamps due to time zones or system clock mismatches
Reconciling data discrepancies between source systems and reported metrics
Documenting all data transformations and cleaning rules for auditability
Validating data completeness against expected sample size and time period
Flagging data quality issues that may require revisiting the collection plan

Module 6: Integrating Data into Process Baseline Calculations

Selecting appropriate metrics: DPMO, sigma level, process yield, or cycle time
Calculating short-term vs. long-term process capability using correct standard deviation formulas
Adjusting for process shifts (e.g., 1.5 sigma) only when justified by historical behavior
Mapping process steps to collected data to compute rolled throughput yield (RTY)
Handling non-normal data using transformations or non-parametric methods
Visualizing baseline performance with time series plots and control charts
Identifying data segmentation opportunities (e.g., by shift, machine, location) for deeper insight
Presenting baseline metrics with confidence intervals to reflect sampling uncertainty

Module 7: Maintaining Data Integrity During Analyze and Improve Phases

Preserving original data sets while creating analysis-specific subsets
Tracking changes to data or assumptions during root cause validation
Collecting additional data to test hypotheses identified in the Analyze phase
Using before/after paired sampling when evaluating solution impact
Ensuring consistency in measurement methods pre- and post-improvement
Documenting data sources and transformations used in statistical models (e.g., regression, ANOVA)
Validating that new process data reflects sustained changes, not temporary fixes
Archiving raw data, analysis scripts, and output for future replication

Module 8: Sustaining Data Collection in the Control Phase

Transitioning project data collection to operational owners with documented handover
Embedding key metrics into existing dashboards or performance reporting systems
Establishing control charts with appropriate control limits and response protocols
Defining frequency and responsibility for ongoing data review and escalation
Updating data collection procedures in standard operating instructions (SOPs)
Training process owners on interpreting control signals and taking corrective action
Conducting periodic audits of data collection adherence and accuracy
Planning for system changes (e.g., ERP upgrades) that may disrupt data continuity