Nonconcurrent Multiple Baseline Designs in Organizational Behavior Management

Single-case research designs are essential tools for demonstrating whether workplace interventions actually work. But real organizations rarely cooperate with ideal research timelines. This article explores how the nonconcurrent multiple baseline design offers a practical path forward—one that balances scientific rigor with the messy realities of organizational life.

What is the research question being asked and why does it matter?

The core question is straightforward: How do you run a solid single-case study in a workplace when you cannot start all baselines at the same time?

In many job sites, teams have different schedules, leaders, and deadlines. This makes a traditional concurrent multiple baseline difficult—it requires all groups to collect baseline data during the same calendar period.

This matters because OBM teams still need a fair test of whether an intervention caused change. Without a believable cause-and-effect link, you risk rolling out a training, feedback system, or policy change that does not actually help. You also risk blaming staff for changes driven by something else entirely—a new manager, a company-wide policy shift, or seasonal factors.

The paper focuses on the nonconcurrent multiple baseline (NCMBL) design as a practical alternative. Each tier (person, team, location, or process) can start baseline at different times while the intervention is still staggered. The goal is to maintain enough experimental control to reasonably claim that the intervention—not random workplace events—produced the change.

What did the researchers do to answer that question?

This was a methods and practice article, not a new treatment study. The authors described how NCMBL designs work and why they fit common OBM constraints like uneven schedules and coordination challenges across sites. They compared NCMBL designs to group comparisons and standard single-case designs, weighing what you gain against what you give up.

They also addressed the biggest threat to trustworthiness: history effects. A history effect occurs when something outside your plan changes behavior—a new corporate incentive, a safety rule, a supervisor change, or a major staffing shift. Because NCMBL tiers happen at different times, history effects can mimic intervention effects.

To strengthen the design, the authors described ways to increase confidence that history is not driving outcomes. Replication is key: showing the same pattern of change when the intervention is introduced across two or more tiers. They also noted that NCMBL designs can be strengthened by mixing in other single-case features when feasible—like adding a withdrawal within tiers or using multielement comparisons.

They provided a simple workplace example across three company locations. Each site started baseline on different dates, and each improved only after performance feedback began at that site. The example demonstrated how you can display the data clearly even when tiers do not align in time.

How you can use this in your day-to-day clinical practice

If you work in OBM and cannot get all teams to start baseline together, an NCMBL can be a solid option. You can still test your intervention carefully without forcing an unrealistic schedule on the organization. This is especially helpful when working across sites, departments, shifts, or teams that cannot “wait” for your study timeline.

The practical shift: you can plan a staggered rollout that matches real constraints while still building a believable case that your intervention helped.

Choose your tiers carefully. Decide what your tiers are and ensure they are meaningfully separate. In OBM, tiers might be different stores, units, classrooms, shifts, teams, or workgroups. They can also be different processes—like steps in a workflow—but only if changing one step does not immediately change the others. The more your tiers affect each other, the harder it is to interpret results. Pick tiers with as little cross-talk as possible.

Plan baseline with flexibility. Treat baseline like you would for any multiple baseline, but accept that start dates can differ. You still need repeated measurement and a clear behavior definition. In workplaces, good targets are usually direct and countable: number of completed checklists, percent of tasks done correctly, rate of safety gear use, or time to close a ticket. Keep measurement steady across time so you do not accidentally change the data system when you introduce the intervention.

Track history effects deliberately. In OBM, history is not rare—it is normal. Track it instead of pretending it will not happen. Keep a simple event log: policy changes, new supervisors, staffing shortages, pay changes, new software, seasonal workload, major incidents. When reviewing graphs, check whether those events align with behavior changes. Be cautious about crediting the intervention if the timing does not fit.

Use the staggered rollout strategically. The stagger protects against history, not just scheduling convenience. If you introduce the intervention in Tier 1 and behavior changes, wait to see whether Tier 2 stays in its baseline pattern before starting there. If Tier 2 changes before you introduce the intervention, treat that as a warning sign. You may need to pause, extend baseline, adjust the plan, or select different tiers.

Build in enough tiers. More replications across tiers reduce the chance that a single history event explains everything. A two-tier NCMBL is weaker than a three- or four-tier design, especially in unstable workplaces. When you only have one or two tiers available, treat results as more tentative and rely more on other evidence—procedural fidelity checks and stakeholder feedback about what actually changed.

Be honest about limits. An NCMBL can help you show that behavior tends to change after the intervention starts, across tiers, even when baseline start times differ. It cannot fully rule out every history event, and it does not automatically generalize to other sites or job roles. If you are using findings to guide a larger rollout, treat it like a strong pilot and keep monitoring outcomes after scale-up.

Strengthen the design within tiers when possible. For some OBM interventions, removing the intervention to run a withdrawal may not be practical or ethical—especially if it affects safety or required performance standards. If you cannot withdraw, you can still improve confidence by tightening fidelity, keeping measurement constant, and showing clear, repeated changes across tiers at the point of intervention.

Plan visuals for stakeholders. With NCMBL, tiers do not line up in time, so make sure dates are clear and phase change lines are obvious. If leaders might misread stacked graphs, consider a timeline view showing when each baseline and intervention occurred in real calendar time. The goal is to reduce the chance that decision-makers over-claim the effect or miss a possible history issue.

Use NCMBL as a decision tool. If your data show change only after feedback starts across several teams, that supports continuing or expanding the approach—while still watching for context changes. If results are mixed, the design helps you troubleshoot by asking, “What was different in this tier?” rather than assuming staff did not buy in. This keeps the focus on fit, support, and dignity instead of blame, and helps you make smarter next steps based on what the data actually show.

Works Cited

Harvey, M. T., & Kennedy, C. H. (2025). Nonconcurrent multiple baseline designs for applied research in organizational behavior management. Journal of Organizational Behavior Management. https://doi.org/10.1080/01608061.2025.2581028