Introduction to the Confidence Threshold App

In monitoring and evaluation (M&E), decisions are rarely made with perfect information. Evaluators must often determine whether the available data provides enough certainty to support a conclusion or recommendation. The Confidence Threshold App is an interactive tool designed to help practitioners determine the appropriate confidence score threshold based on the level of risk associated with a decision.

By aligning the required level of confidence with the stakes of a decision, the app supports more transparent, consistent, and defensible evaluation practices. Whether you are conducting a programme review, assessing humanitarian funding, or collecting formative feedback, the tool helps clarify how much certainty is needed before drawing conclusions.

Why Confidence Thresholds Matter

Different evaluation tasks require different levels of certainty. For example, a decision that affects national funding allocations should require much stronger evidence than a workshop feedback survey. Setting the right confidence threshold helps balance two important considerations:

• Decision risk – the consequences of being wrong.
• Evidence quality – how reliable the available data is.
• Accountability – ensuring decisions are transparent and defensible.
• Efficiency – avoiding unnecessary analysis when the stakes are low.

How the Confidence Threshold App Works

The tool provides an interactive interface where users can select typical monitoring and evaluation scenarios or manually adjust the risk level of a decision. Based on the selected context, the app suggests a recommended confidence threshold.

Users can explore scenarios such as:

• Humanitarian funding allocation decisions
• Programme certification audits
• Annual performance reviews
• Mid-term programme evaluations
• Internal learning or formative feedback exercises

The app then calculates a recommended confidence level — typically ranging between 60% and 95% — depending on the risk and consequences of the decision.

Typical Threshold Guidance

The tool is based on a simple risk-based framework commonly used in evaluation and evidence assessment:

• High-stakes decisions: ≥ 90% confidence
• Medium-stakes decisions: 75–89% confidence
• Low-stakes decisions: 60–74% confidence

These thresholds provide practical guidance for balancing statistical confidence, qualitative evidence, and evaluation judgement.

Supporting Better Evaluation Decisions

By explicitly linking evidence confidence with decision risk, the Confidence Threshold App helps evaluation professionals:

• Standardize evidence interpretation
• Improve transparency in evaluation reporting
• Strengthen methodological justification
• Facilitate discussions with stakeholders and decision makers

The tool is particularly useful for monitoring and evaluation specialists, research teams, development organizations, and policy analysts working in complex decision environments.