Earned Value Definitions
Variance Threshold

What is Variance Threshold? Earned Value Management explained.

Mia Rutherford
Mia Rutherford
February 24, 2024
5 min read

Variance Threshold is a pivotal concept in project management, particularly within the framework of Earned Value Management (EVM). It helps in determining the significance of variance in project performance measures. By understanding and applying the principles of Variance Threshold, project managers can effectively monitor and control project progress, ensuring that project objectives are met within the constraints of time and cost. This article delves into the intricacies of Variance Threshold, offering insights into its calculation, estimation techniques, and practical applications in various case studies.

Key Takeaways

Understanding Variance Threshold in Project Management

Defining Variance Threshold in the Context of EVM

In the realm of project management, Earned Value Management (EVM) stands as a pivotal methodology for measuring project performance and progress. A key component within this system is the variance threshold, which serves as a benchmark for identifying significant deviations from the project plan.

The variance threshold is essentially a predetermined point at which a variance is considered significant enough to warrant attention and, potentially, corrective action. This concept is crucial because it helps project managers to:

Variance analysis, as highlighted by Simplilearn.com, is a fundamental practise in EVM that involves comparing planned values (PV) and earned values (EV) with actual costs (AC). By calculating these variances, project managers can determine whether they are over or under budget and schedule, and by how much. The variance threshold then acts as a guide to gauge the severity of these variances and decide on the necessary course of action.

The Role of Variance Threshold in Project Performance

In the realm of project management, the variance threshold serves as a pivotal metric for gauging project health and steering projects towards successful completion. It acts as a benchmark for acceptable deviations from planned values, whether they pertain to cost, time, or scope. By setting a predefined variance threshold, project managers can swiftly identify when a project is veering off course and implement corrective actions before minor issues escalate into significant setbacks.

The practical implications of an effective variance threshold are manifold:

Ultimately, the variance threshold is not just a number; it encapsulates the project's tolerance for risk and uncertainty. Establishing a robust variance threshold is a testament to an organisation's commitment to diligent project management and cost efficiency. As such, it is a cornerstone in the edifice of Earned Value Management (EVM), ensuring that projects are not only completed within their set parameters but also deliver the expected value to stakeholders.

Calculating Variance Threshold: A Step-by-Step Guide

The process of calculating the variance threshold is a meticulous one, requiring a clear understanding of the project's financial metrics. Earned Value Analysis (EVA) is crucial in project management for cost forecasting, performance evaluation, and effective decision-making. It involves key metrics like PV, EV, AC, SV, CV, SPI, and CPI.

To begin with, one must first establish the baseline for the project, which is the approved budget and schedule. Following this, the actual cost (AC) and the earned value (EV) are calculated. The variance threshold is then determined by comparing these figures against the planned value (PV). Here's a simplified step-by-step approach:

  1. Calculate the Planned Value (PV).
  2. Determine the Earned Value (EV).
  3. Ascertain the Actual Cost (AC).
  4. Compute the Schedule Variance (SV) and Cost Variance (CV).
  5. Establish the Variance Threshold based on the project's tolerance levels.

It's essential to note that the variance threshold is not a static figure; it should be reviewed and adjusted as the project progresses to reflect any changes in scope, schedule, or cost. This dynamic approach ensures that the project remains on track and within the acceptable limits of variance, thereby safeguarding the project's financial health and success.

The Mechanics of Variance Estimation

Patient-by-Patient Variance Estimation Techniques

In the realm of project management, particularly within Earned Value Management (EVM), the ability to estimate variance on a patient-by-patient basis is crucial for maintaining control over project costs and timelines. This granular approach allows for the identification of specific areas where resources may be over or underutilised, leading to more accurate forecasting and budgeting.

To effectively implement patient-by-patient variance estimation, one might consider the following steps:

  1. Collect detailed data on each patient's costs and progress.
  2. Analyse the data to identify any significant deviations from the planned values.
  3. Apply statistical methods, such as standard deviation and range, to quantify the variance.
  4. Investigate the root causes of any identified variances to inform corrective actions.

By adopting these techniques, project managers can enhance their ability to spot patterns and trends, which is essential for proactive project management. For instance, employing a Pareto Analysis can help in prioritising the most significant cost variances and applying the 80/20 rule to focus on the areas that will have the greatest impact on project performance.

Table 5: Interpreting the Common Within-Patient Variance

Interpreting common within-patient variance is a critical step in understanding the nuances of patient-specific responses in medical studies. Table 5 provides a comprehensive overview of this variance, allowing practitioners to discern patterns and anomalies in patient data. This interpretation is pivotal for tailoring treatments and predicting outcomes with greater accuracy.

The table typically includes statistical measures such as the mean, standard deviation, and the ANOVA results. These figures are instrumental in estimating the variability of a patient's response over time. For instance:

By mastering the interpretation of such data, medical professionals can enhance their diagnostic precision and optimise therapeutic strategies. It is also a valuable skill in the broader context of Earned Value Management (EVM), where understanding variance is essential for successful project delivery.

Fig. 4: Pooling Prior Variance with Current Patient Data

In the realm of Earned Value Management (EVM), the technique of pooling variances is a pivotal analytical tool. It allows project managers to integrate historical data with current performance metrics to form a more comprehensive understanding of project variability. Fig. 4 illustrates the technique of pooling a prior variance with the variance from a given patient, in this case, patient number 5. This approach is particularly useful when dealing with continuous data from n-of-1 trials, where each patient's data is treated as a single trial.

The process involves several steps:

  1. Identifying the prior variance from historical data or previous trials.
  2. Calculating the current patient's variance based on new data.
  3. Combining these two variances to estimate a pooled variance.
  4. Establishing 95% confidence limits around the pooled estimate to assess the precision of the estimate.

By adopting this method, project managers can make more informed decisions, as the pooled variance provides a more robust estimate of the true variability within the project. It also aids in identifying any outliers or trends that may not be apparent from a single set of data. The analysis of continuous data from n-of-1 trials using paired cycles becomes more reliable, and the resulting insights can lead to improved project outcomes.

Incorporating External Estimates in Variance Analysis

Utilising External Variance Estimates in EVM

In the realm of Earned Value Management (EVM), the integration of external variance estimates can be a game-changer for project managers seeking to enhance accuracy in forecasting and risk assessment. By incorporating variance data from similar projects or industry benchmarks, project managers can calibrate their own variance thresholds with greater precision.

When external estimates are employed, it is crucial to ensure that they are relevant and reflect the current market conditions. This can involve a series of steps:

  1. Identifying credible sources of external variance data.
  2. Adjusting the data to align with the specific context of the project.
  3. Integrating the adjusted data into the EVM system to refine the variance threshold.

This approach not only enriches the project's data pool but also provides a more robust framework for decision-making. It allows for a proactive response to potential cost overruns and schedule delays, ultimately contributing to the project's success.

Post-hoc ANOVA Tests and Pooled Variances

Upon the completion of an ANOVA, which discerns whether there are significant differences across group means, post-hoc tests are employed to pinpoint exactly where these differences lie. These tests are crucial in environmental and biological sciences, as they allow for a detailed comparison between individual group means after a significant variance has been established.

The process of pooling variances is another essential step in the analysis. It involves combining the within-group variances to create a single, more stable estimate of variability. This pooled variance is then used in subsequent post-hoc tests to ensure a more accurate and reliable comparison between groups.

Here are the key steps in conducting post-hoc ANOVA tests with pooled variances:

  1. Perform the initial ANOVA to check for overall significance.
  2. Calculate the pooled variance from the within-group variances.
  3. Apply the post-hoc test using the pooled variance to determine specific group differences.
  4. Interpret the results to understand the practical implications for project management.

Fixed and Random Effect Approaches in Variance Analysis

In the realm of Earned Value Management (EVM), understanding the nuances of variance analysis is crucial for accurate project forecasting and control. Fixed effect models assume that individual project elements have specific, unchanging effects on performance metrics. These models are often used when the project environment is stable and predictable.

Conversely, random effect models account for the inherent variability in project elements, acknowledging that these effects can fluctuate over time. This approach is particularly beneficial in complex projects where uncertainty is a significant factor. By incorporating random effects, analysts can obtain a more realistic picture of project performance, leading to better decision-making.

When applying these models in variance analysis, consider the following steps:

  1. Identify the key project variables that may have fixed or random effects.
  2. Determine the appropriate model based on the project's complexity and the nature of the variables.
  3. Apply the chosen model to analyse the variance and interpret the results to guide project adjustments.

The selection between fixed and random effect models can profoundly influence the accuracy of variance analysis, ultimately impacting project success. It is essential to weigh the characteristics of the project against the strengths and limitations of each model to make an informed choice.

Advanced Concepts in Variance Threshold

Assumptions and Implications of Constant Within-Patient Variation

In the realm of Earned Value Management (EVM), the assumption of constant within-patient variation is pivotal for accurate variance analysis. This presupposes that the variability in performance metrics is uniform across different time periods for the same patient or project component. However, this may not always hold true in practise, leading to potential discrepancies in the variance threshold calculations.

The implications of this assumption are far-reaching. For instance, if the actual variation is higher than anticipated, it could signal underlying issues such as inadequate data collection or changes in project scope that have not been accounted for. Conversely, lower than expected variation might indicate a well-controlled process, but could also mask emerging problems that are not yet reflected in the data.

To mitigate these risks, project managers should:

Formula for the Variance of the Shrunk Estimate

The concept of a 'shrunk' estimate is pivotal in the realm of variance analysis, particularly when dealing with n-of-1 trials. The essence of this estimate is the reduction in variance that it brings, ensuring a more precise and reliable outcome. The formula for the variance of the shrunk estimate is given by:


This mathematical expression encapsulates the balance between the personal and global estimates. The 'shrunk' estimate is a weighted average of these two, where the weight (w) is inversely proportional to their respective variances:

$$shrunk=w\times personal+\left(1-w\right) global.$$

To elucidate further:

The weighting factor is crucial as it determines the extent to which each estimate influences the final 'shrunk' estimate. It is important to note that the variance of the shrunk estimate will always lie between the global and personal variances, offering a more nuanced and potentially more accurate reflexion of the true variance.

In scenarios where local information is absent (ki = 0), the variance formula simplifies to ψ2, relying solely on global information. Conversely, as ψ2 approaches infinity, the formula converges to the personal variance, indicating that other patients' results do not contribute any information. The beauty of this approach lies in its flexibility and adaptability to the amount of information available, making it a powerful tool in the analysis of continuous data from n-of-1 trials.

Guidance for Designing N-of-1 Trials: The DIAMOND Project

The DIAMOND Project provides a comprehensive framework for designing N-of-1 trials, which are single-patient studies used to determine the efficacy of a treatment for an individual. These trials are particularly valuable in assessing patient-specific responses to interventions, allowing for a more personalised approach to healthcare.

Key considerations in the design of N-of-1 trials include the selection of appropriate outcome measures, the determination of the number of treatment periods, and the use of suitable statistical methods to analyse the data. The DIAMOND Project emphasises the importance of:

By adhering to these guidelines, researchers can maximise the reliability and validity of their findings, ultimately contributing to more effective and individualised patient care.

Practical Applications and Case Studies

Case Study Analysis: Applying Variance Threshold in Real-World Scenarios

The practical application of variance threshold in project management can be transformative, particularly when integrated with tools like site diaries and Earned Value Metrics. By examining real-world scenarios, we can discern how meticulous tracking and analysis of variance can preemptively identify potential project pitfalls.

For instance, in construction management, the use of site diaries for real-time reporting allows for the immediate capture of discrepancies between planned and actual progress. This proactive approach not only facilitates timely corrective actions but also serves as a robust defence mechanism against disputes. By setting a variance threshold, project managers can establish clear benchmarks for when to intervene, ensuring that projects remain on track and within budget.

Moreover, the impact of variance threshold extends to enhancing communication among stakeholders. When variance exceeds the predefined threshold, it triggers a review process, prompting discussions that can lead to strategic adjustments. This level of transparency and responsiveness is crucial for maintaining project momentum and stakeholder confidence.

Software Tools for Variance Analysis: Genstat for Windows

In the realm of variance analysis, software tools play a pivotal role in simplifying complex calculations and visualising data trends. Genstat for Windows stands out as a comprehensive tool designed for professionals in construction and project management. It offers a suite of features tailored to enhance the accuracy and efficiency of variance analysis.

Key functionalities of Genstat include:

By leveraging Genstat, project managers can streamline project data analysis, automate progress tracking, and improve cost management. This facilitates collaborative decision-making and ensures that projects remain on track. The software's ability to handle large datasets with precision makes it an invaluable asset for those seeking to maintain stringent control over project variables and outcomes.

Evaluating the Impact of Variance Threshold on Project Outcomes

The implementation of a variance threshold within Earned Value Management (EVM) systems can be a game-changer for project outcomes. By setting a predefined limit on acceptable variances, project managers can proactively identify and address deviations from the project plan. This vigilance ensures that projects remain on track, both in terms of budget and schedule, and can significantly enhance the overall success rate.

To evaluate the impact of variance thresholds, one must consider several key factors:

Incorporating variance thresholds into project management practises not only aids in maintaining control over project trajectories but also contributes to effective fatigue management in construction. By preventing overruns and the need for last-minute scrambles, teams can work within a more structured and less stressful environment. Tools like Site Diary and industry standards play a pivotal role in monitoring and mitigating fatigue risks, ensuring worker safety and productivity are upheld.


In summary, the Variance Threshold technique within Earned Value Management is a robust statistical tool that aids in the reduction of variance by constructing patient-specific estimates. As we have explored, this method involves pooling variances and leveraging both fixed and random effect approaches to yield more precise estimates. The technique's efficacy is underscored by its ability to maintain a constant within-patient variation across cycles, as well as its adaptability in incorporating external variance estimates in a manner akin to post-hoc ANOVA tests. The practical application of these concepts is exemplified in the detailed calculations and illustrations provided, such as those in Table 5 and Figure 4. It is our hope that this article has elucidated the intricacies of Variance Threshold and its pivotal role in enhancing the accuracy of project management forecasts, ultimately contributing to more informed decision-making processes.

Frequently Asked Questions

What is a Variance Threshold in the context of Earned Value Management?

In Earned Value Management (EVM), a Variance Threshold is a predefined limit that indicates an acceptable range of deviation from the project's planned value and schedule. If the actual performance exceeds this threshold, it triggers a review or corrective action.

How is the Variance Threshold calculated in EVM?

The Variance Threshold is typically calculated based on historical performance data, project complexity, and stakeholder risk tolerance. It involves determining the acceptable percentage of deviation from the planned cost or schedule.

What is the significance of patient-by-patient variance estimation in clinical trials?

Patient-by-patient variance estimation is crucial in clinical trials to account for individual differences in response to treatment. This helps to identify the common within-patient variance and improve the accuracy of the overall variance estimate.

How does pooling prior variance with current patient data affect the analysis?

Pooling prior variance with current patient data allows the incorporation of historical information to refine the estimate of variance for a current patient. This can result in more precise confidence limits and a better understanding of the treatment's effect.

What is the formula for the variance of the shrunk estimate?

The formula for the variance of the shrunk estimate is: Var(shrunk) = (2ψ^2σ^2) / (kiψ^2 + 2σ^2), where ψ represents the prior variance, σ is the current variance, and ki is the number of prior degrees of freedom.

Can external variance estimates be used in EVM, and how do they influence the analysis?

Yes, external variance estimates can be used in EVM to supplement internal data. This approach is in line with post-hoc ANOVA tests, where pooled variances across different segments are common. It helps to enhance the robustness of the variance analysis.


Keep up with Gather

Make sure you never miss out! Sign up to our monthly newsletter to keep up with the biggest news stories in construction and the latest Gather updates. Full of our latest case studies, blogs and fun quizzes!

Thank you for subscribing!
Oops! Something went wrong while submitting the form.

Gather needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at anytime. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, check out our Privacy Policy.