Multi-Subject Studies: Observing Complete Teams

What Are Multi-Subject Studies in Work Sampling?

Multi-subject studies represent a sophisticated evolution of the classic Work Sampling. This technique allows diagnosing the productivity of complete teams simultaneously and objectively. Unlike traditional individual observation, here the state of multiple operators is captured at random instants, known as Snap Readings.

The result is a systemic view of group performance. An accurate X-ray is obtained of how work flows in an entire area, without the need for invasive hardware or interruptions to production. It is a living methodology that has evolved to respond to the challenges of modern manufacturing.

Statistical Foundations for Team Observation

The power of this technique lies in its statistical rigor. It is not about observing arbitrarily, but about applying statistical inference so that the results are meaningful and reliable.

Rigorous Sample Design for Group Inference

It all begins with precise calculation. The validity of multi-subject studies depends on the sample size (N), determined through the binomial distribution. This formula considers two pillars: the desired confidence level (Z) and margin of error.

For a study with a 95% confidence level (Z=1.96) and a margin of error of 3%, the formula is clear. If we work with a team of 10 people and an expected activity proportion of 50% (the maximum variability scenario), approximately 1,067 observations per subject are required. That translates to nearly 10,670 total observations for the team. This rigor is what separates an accurate diagnosis from a simple subjective impression.

Tippett Technique and Snap Reading for Simultaneous Capture

The generation of these thousands of observations is not chaotic. The Tippett technique is used to create random snapshots. An algorithm generates the exact moments to perform a Snap Reading, a quick reading of the state of the entire team.

This strict randomization method is crucial. It minimizes two enemies of objectivity: temporal correlation bias (always observing at the same cycles) and the Hawthorne Effect (workers modifying their behavior when they know they are being observed). The Cronometras tool has greatly simplified the management of these random cycles, allowing the analyst to focus on data capture.

Practical Methodology in Spanish Industrial Environments

Implementing these studies in Spain requires knowing the current regulatory framework. It is not only about efficiency, but also about compliance and respect for labor rights.

Implementation in Compliance with 2025 Regulations

Multi-subject studies must align with several key regulations:

• Occupational Risk Prevention Law (LPRL): Observations must never interfere with safety or generate additional stress.
• General Data Protection Regulation (GDPR): Individual productivity data must be anonymized at the team level. The group is analyzed, not the individual pursued.
• Collective Bargaining Agreements: Many establish limits on continuous monitoring, which fits perfectly with the discrete and random nature of sampling.

MECE Taxonomy for Activity Categorization

For the analysis to be objective, a common language is needed. Here enters the MECE taxonomy (Mutually Exclusive, Collectively Exhaustive). It consists of creating activity categories that do not overlap and that, together, cover 100% of observable time.

A practical example aligned with ISO 9001:2015 would be:

1. Direct Work: Operation, assembly, quality verification (value added).
2. Indirect Work: Internal transport, obtaining materials.
3. Waits: Due to stockout, breakdown, or instructions.
4. Management: Brief meetings, administrative records.
5. Unproductive: Unjustified absences, delays.

This structure ensures that each Snap Reading is classified unambiguously, generating clean data for analysis.

Applications and Key Metrics

The information obtained allows calculating strategically high-impact metrics, traditionally linked to costly technological systems.

OEE Calculation Without Sensors and Wrench Time

OEE (Overall Equipment Effectiveness) usually requires sensors on machines. However, through multi-subject observations, we can estimate OEE without sensors. By classifying waiting times and machine adjustments, availability and performance losses are reliably calculated.

Another critical metric is Wrench Time (active tool time). It is the percentage of time the operator dedicates to their main value-added task. Multi-subject studies are the standard method to measure it. For real-time production control, platforms like Induly offer the possibility of crossing these observation data with production records, validating the findings.

Empirical Case: Spanish Automotive Plant

Data confirms it. In a study conducted at an automotive plant, 12 operators were observed for 5 working days. 480 random readings were generated (10 readings/hour, 8 hours/day).

The results were revealing:

• Team average Wrench Time: 58% (with a 95% Confidence Interval and a margin of error of ±3.2%).
• Activity distribution:
  • Direct work (value added): 58%
  • Waits and transfers: 22%
  • Machine adjustments: 12%
  • Administrative time: 8%

This diagnosis made it possible to identify that almost a quarter of the time was lost in waits and transfers, a clear opportunity to optimize internal logistics.

Software and Tools for Multi-Subject Studies

Current technology enhances this classic methodology, making it more agile and precise.

Specialized Technical Solutions

Modern software for multi-subject Work Sampling should offer:

• Simultaneous registration of multiple subjects (up to 20 per observer).
• Automatic generation of routes and Snap Reading alerts based on Tippett.
• Real-time calculations of remaining sample size, confidence intervals, and margin of error.
• Direct export to dashboards for metrics like OEE and Wrench Time.

Work sampling, facilitated by applications like WorkSamp, allows implementing these studies with a rigor previously only within reach of large consulting firms.

Integration with the Plant's Digital Ecosystem

The value of data multiplies when integrated. Work Sampling software must be compatible with:

• ERP (SAP, Oracle): To cross time data with costs and planning.
• CMMS (Maintenance Management): To correlate observed machine stops with work orders.
• Simulation tools: To model the impact of improvements before implementing them.

Resources and Tools

To delve deeper into these methodologies and find the right technology partners, we recommend the following industry resources:

• ASETEMYT: The reference directory for finding specialists in timing, methods, and work study. Explore its professional directory.
• WorkSamp: Specialists in Work Sampling. Their platform is designed specifically for the rigorous implementation of Work Sampling studies, including multi-subject ones.
• Cronometras: Essential tool for time and motion analysis, complementary to sampling for full cycle studies.
• Induly: Production Control and Industrial Time-Clocking Software. Ideal for monitoring productivity in real time and validating data obtained by observation.
• ASETEMYT Blog: Technical articles and case studies on methods engineering and productivity.
• Add your company to the directory: If you offer timing services or software, you can be part of the ASETEMYT ecosystem.

Multi-subject studies are not a relic of the past, but the quantitative foundation for decision-making in the productivity of the future. They combine classic statistical rigor with current digital tools, offering a reliable, objective, and legally compliant diagnosis. It is the essential starting point for any serious continuous improvement initiative.