What is Control Chart in PMP? A Complete Guide with Examples

As a control chart has allowed me to improve constantly over the past ten years, I find it hard to select any particular tool from the arsenal of quality management. The first time I stumbled across control charts whilst preparing for my PMP certification, I did not truly believe in their value. Now, I can confidently say that I have never been able to run a successful project without using control charts.

In this guide, I will give you fundamental starting concepts that can aid you in mastering control charts in project management, including lessons where you can leverage them in passing your PMP exams and, even more significantly, in executing projects successfully.

Control Charts have extensive uses in Project Management, especially in distinguishing between the normal fluctuations due to common causes and the uncommon variations that need immediate rectification, also known as Special Causes. Control charts were created by Bell Laboratories' physicist and engineer Walter Shewhart in the 1920s as part of statistical process control (SPC) systems.

The "Control Quality" process in the Monitoring and Controlling process group of the PMP framework includes control charts. As part of data-driven project management, control charts enable project managers to:

• Observe the behavior of processes.
• Decide if a process can be identified as stable and predictable.
• Decide when a process requires corrective action.
• Ensure that designated improvements have been actualized.
• Assist in reporting the results of a project, along with verifying the project's efficiency regarding the communicated scope and stakeholder needs.

As a novice, one of my most memorable screw-ups was while working on a software development project where we were constantly and mindlessly deploying new features. This ultimately resulted in the gradual degradation of our system's performance. One unfortunate day, we suffered a major outage. To avoid issues, post the incident, I started implementing control charts to measure performance metrics so we could identify troubling code changes before they were delivered to users.

Control charts automated project management, and shifted the focus from reactive to proactive. Instead of waiting for scenarios to unfold, managers can anticipate developing trends well ahead of time and intervene as necessary.

This is why control charts are helpful for project managers in the building:

• Early Warning System: They notify you about deviations needing attention prior to complications arising.
• Objective Decision Making: They offer statistical data in support of mitigations that lessen subjective influence.
• Process Improvement: They serve to determine the capability and stability of the processes.
• Regulatory Compliance: They furnish documentation of process control that may be required in regulated industries.
• Team Accountability: They ensure visibility into performance, thus aiding the motivation of the teams in adhering to standards.

To make effective use of control charts in project management, one needs to know the components. These include:

• Centerline (CL): Indicates the average of the data points
• Upper Control Limit (UCL): Generally preset at 3 standard deviations above the mean
• Lower Control Limit (LCL): Generally preset at 3 standard deviations below the mean
• Data Points: Sequences of measurements shown graphically in time order
• Time Axis: Denotes the period within which the provided measurements were taken
• Measurement Axis: Indicates the measurement value for each plotted sequence

Let us further reveal the significance of these components. In the case of statistical process control (SPC), the major components are control limits, which include both upper control limit (UCL) and lower control limit (LCL), which define the range of fluctuation that is often anticipated in a process that is stable.

If data points fit within these boundaries and don't display patterns, the process is identified as "in control." This does not indicate that the process is good or fulfills requirements; it simply shows that it is stable and predictable.

All control charts are not created equal. The type of control chart to use is determined by the data being collected and what the user aims to measure. Here is an extensive summary incorporating the common types:

Variable Data Control Charts

These charts track characteristics that can be measured on a continuous scale, time, weight, or temperature.

1.X-bar and R charts: Used when there is a possibility of getting samples of 2–10 observations with time intervals.

• X-bar is for sample means
• R is for sample ranges

2.X-bar and S charts: Similar to X-bar and R charts. Instead of the range, the standard deviation is used.

• Better for larger subgroups (>10)
• More sensitive to variation

Individual and Moving Range (I-MR) charts: Used for situations where data is collected one observation at a time.

• Each value can be tracked in the individual chart
• The moving range chart is responsible for tracking the difference between the preceding and succeeding values.

Attribute Data Control Charts

These charts monitor characteristics that are counted, not measured, such as defects or results that are passed or failed.

• The proportion of defective items in p charts can be tracked in samples of varying sizes.
• In samples of constant size, the number of defective items in np charts can be tracked.
• For c charts, the number of constant-size defects in the sample is tracked.
• In u charts, the number of defects per unit in samples of varying sizes is tracked.

Below is a comparison table designed to aid your selection of the most effective control chart for your project requirements:

To be honest, when I created my first control chart, all the statistical formulas made it look so complicated.

As I put in effort, it started coming naturally. This is what I did:

1. Identify What Requires Measurement

Identify process features that are important for:

• Enhancing customer satisfaction.
• Fulfilling project success criteria.
• Previously have presented issues.
• Compliance prerequisites.

2. Data Collection

For proper control charts:

• A minimum of 20 to 25 measurement points must be collected.
• Single measurement means must be put in place.
• Recorded time periods of data collection.
• Subjective Controversial situations.

3. Establishing Control Limits

For an I-MR control chart (which is one of the most straightforward to determine):

• Determine the moving range average (X)
• Determine the moving ranges (MR) of a dataset between two contiguous points.
• Accomplish the average moving ranges (MR ring)
• Set UCL = X̄ + 2.66 × MR̄.
• Set LCL = X̄ - 2.66 × MR̄

4. Draw the Chart

• Insert the overall average's centerline.
• Illustrate UCL alongside LCL.
• Plot chronological sequence data points.
• Draw lines between connecting and consecutive points.

5. Evaluation

Perform close searches for:

• Independently controlled points.
• Ongoing progression for 7 + a subset of mid-range points.
• Trends of 7+ from a single mid-range point consistently increasing or dipping downwards.
• Apply a diagnosis or some randomly structured patterns.

6. Responding to Appropriate Action Take:

• When processes are under control, observing and responding to any necessary capability improvements is needed.
• For uncontrolled processes: Examine unique causes, take actions

For me, creating control charts is possible with Minitab and SPC XL, and even Excel. Although Excel control charts require a higher degree of effort, it is available on all computers, which is beneficial for elementary control charts.

Recognizing patterns is critical for effective control chart analysis within the scope of PMP practice. Here are details of major patterns to keep in mind.

1. Points Beyond Control Limits

• Data points lying above UCL or below LCL.
• A process is being influenced by a special cause; therefore, further analysis is required without delay.
• Example: In a call center, a sudden spike in call duration might indicate a system outage.

2. Trends

• 7 or more data points, either increasing or decreasing in a stepwise fashion.
• Gradual change is impacting the process; it could be deterioration, improvement, or seasonal.
• Example: Gradually increasing error rates might indicate equipment wear-out.

3. Shifts

• Seven or more data points in a row sequentially above and below the centerline of the chart.
• The process average has been altered by something.
• Example: After training, all points below the centerline for processing time indicate improved efficiency.

4. Cycles

• More than two occasions where a series of values displays clear, predictable up and down movements.
• Processes are impacted by factors such as staff rotation, maintenance schedule, or other cyclically recurring phenomena.
• Example: Higher defect rates every Monday may indicate issues with weekend-to-weekday transitions.

5. Hugging the Centerline

• What it looks like: Most or all points are clustered around the centerline. Few points approach the control limits.
• What it means: The observation may result from blending distinct data sets or calculation error. Alternatively, the system used may not have sufficient granularity or precision.

In my estimation, the ability to "read" control charts is what differentiates the novice from the expert in project management.

Here are some real-life examples of control charts in a PMP context that I have done:

Example 1: Software Development Bug Tracking

In one of the software development projects, we tracked how many bugs were found per 1,000 lines of code around sprints. Each time we looked at a new Sprint, we used a U-chart as the 'samples' were not of uniform size.

Having created our control charts, we saw that Sprint 7 was spiking disproportionately in terms of bugs. Upon further investigation, we discovered that there was a new developer who was onboarded and severely trained on the set coding standards. After some remediation, normal variation was observed.

Example 2: Construction Project Schedule Variance

We monitored the schedule variance percentage on a weekly basis using an I-MR chart:

• Week 1: 2.1%
• Week 2: 1.7%
• Week 3: 2.2%
• Week 5: 1.8%
• Week 6: 2.3%
• Week 7: 2.6%
• Week 9: 3.2%
• Week 10: 3.4%
• Week 11: 3.5%
• Week 12: 4.1%

Control charts indicated the start of increasing variance around week 8. Inspecting this change revealed that a critical supplier was slowly lagging on all their deliveries. Actively managing this concern early on helped save substantial schedule slippage later on.

Throughout the years, I have compiled these best practices for implementing control charts that have worked well for me:

• Set project milestones and schedules early on in the lifecycle of the project
• Create baselines during the initial planning phases
• Do not wait for problems to appear
• Define relevant metrics
• Concentrate on the most important critical success factors
• Balance leading indicators and lagging indicators
• Start simple.
• Work collaboratively.
• Ensure all team members understand basic statistics
• Make data gathering part of daily responsibilities
• Have regular conversations and compare shared charts on a scheduled basis

• Select the correct software.

  
  1. Excel suffices for straightforward tasks.
  2. Complex applications require dedicated SPC software.
  3. Integrate with project management applications.

   

• Maintain uniformity.

  
  1. Standardized methods of collecting data must be adhered to.
  2. Control limits must be calculated properly.
  3. Charts must be updated regularly, but limits should not be recalculated too often.

   

• Learn and act from what the data shows.

  
  1. Identify special causes immediately after identifying them.
  2. They must also be documented with actions taken.
  3. Changes to processes over time need monitoring.

   

• Evolve your approach.
   

1. This is done by starting with control charts, then sophistication will be added when needed.
2. Check regularly which metrics deliver value and adjust control limits when processes change.

"The goal is not to do statistical quality control. The goal is to provide the tools and strategies to enable us to compete in the world marketplace." — W. Edwards Deming

If you begin preparing for this exam, study in detail control charts as they will surely come up in the exam questions.

Here is what you should focus on regarding my experience and coaching PMP learners:

PMP Exam Question Types:

1.Calculation questions

• Computing control limits
• Determining if a process is in or out of control

2.Application scenarios

• Selecting the appropriate control chart type
• Interpreting patterns
• Deciding when to investigate versus when to leave a process alone

3.Integration with other processes

• How do control charts in PMP relate to other quality management tools
• When to use control charts in the project lifecycle

For me, control charts are among the strongest tools that a project manager can use. They turn quality management from subjective judgment into objective analysis by enabling one to differentiate between normal variation and issues that need fixing.

In my career, I've benefited from control charts in project management by:

• Being more confident and reliant on data for decision making
• Directing teams to address real issues instead of normal variation
• Proving through statistics that processes are improving
• Interacting with stakeholders on quality issues using clear language

Understanding control charts is not just about passing the test for those preparing for the PMP exam, but rather adopting a methodology that will serve them in their career as project managers.

When you start using control charts for your personal projects, simplify the process, stay consistent, and pay attention to the patterns that shape the narrative of your processes. That level of rigor will shift your project management from being reactive, subjective, and moderately effective, to proactive, objective, and outstanding.