Data, Information, Knowledge: The Confusion That Costs Projects Dearly | Delta Monitoring
Data, information, knowledge: three terms used interchangeably in logical frameworks, terms of reference and project reviews. Yet this conflation causes silent but massive damage: inappropriate decisions, unreadable dashboards and lost learning opportunities.
Confusing a physical completion rate (data) with an alert about a supply delay (information), or with a consolidated rule on recurring bottlenecks (knowledge), dooms monitoring and evaluation to remain an administrative function rather than a management tool.
This article shows why this lack of distinction persists, how it manifests in practice within projects, and how to overcome it
Clarifying the three concepts: a non-negotiable prerequisite
Before any analysis, conceptual clarification is essential. In information science, this triptych is fundamental.
From raw data to capitalised knowledge — three levels of abstraction too often confused in M&E systems.
1. Data: the raw, uninterpreted fact
Data is a raw element, often quantitative, devoid of intrinsic meaning.
Examples: 42% · 15/05 · 1,200 beneficiaries.
A single piece of data does not allow for any decision.
Operational properties
- May be collected automatically or manually;
- Needs to be verified (reliability) and contextualised;
- Becomes useful only if it answers a question.
2. Information: data placed in context and made meaningful
Information arises from the interpretation of data. It is based on:
- A comparison (actual vs target, trend vs threshold);
- Context (area, period, sub-group);
- An operational question (“Is this 42% rate good or bad?”)
3. Knowledge: internalised and reusable information
Knowledge emerges when information is:
- Validated by multiple stakeholders (field team, partners, beneficiaries);
- Repeated over several cycles;
- Incorporated into reusable rules, procedures or reasoning.
The three operational pitfalls in projects
In practice, the confusion between data, information and knowledge produces recurring and objectively identifiable symptoms.
1. Knowledge: internalised and reusable information
Many M&E systems continue to operate under the following implicit assumption: the more data we collect, the better our decisions will be.
The consequences are well documented:
- Collection of indicators with no use (‘orphan’ indicators);
- Overburdening of teams;
- Poor use of the indicators produced
2. Dashboards that don't track anything
A dashboard saturated with data becomes unusable. The phenomena observed include:
- Lack of explicit alert thresholds;
- Visualisations that are difficult to use;
- Confusion between activities and results.
3. Lost learning
The most costly pitfall: information correctly identified during a project cycle is never transformed into knowledge for the next cycle or for other projects.
This phenomenon results in:
- Repetition of errors from one phase to the next (same bottlenecks, same delays);
- Reliance on key individuals who are “subject matter experts”;
- Ineffective capitalisation: lessons learnt without a mechanism for reuse.
A project that fails to transform its information into knowledge constantly relearns what it should already have mastered.
Why does this confusion persist?
Three structural biases explain its persistence in development projects.
1. Donor bias: the demand for quantified evidence
The pressure to produce quantified and ‘verifiable’ indicators often pushes teams to prioritise quantity over usefulness. Not all metrics are worth tracking.
2. Tool limitations: databases without a semantic layer
Most M&E tools remain data managers. They do not include the transformation of data into information or knowledge. Without this layer, each user interprets the same data differently.
3. Human bias: confusing intellectual activity with action
Producing figures gives the impression of progress. This feeling often masks a more uncomfortable question:
Does this data change a decision?
The result is cognitive overload coupled with sub-optimal decision-making.
Restoring a functional decision-making chain: three operational principles
Moving from confusion to control requires three simple principles, though they are demanding to implement.
1. All data collected must serve information necessary for a decision
Before adding a variable or indicator, systematically ask two questions:
- Which operational question does this data answer?
- What specific decision will be affected by the answer to this question?
If no decision is identified, there is no point in collecting the data.
2. Explicitly separate the levels in M&E outputs
A good dashboard does not mix three levels in the same view. Instead, it offers:
- A raw data section (verifiable, accessible);
- An analysed information section (alerts, deviations, trends, with thresholds);
- A consolidated knowledge section (recap of previous learnings, recommendations from previous cycles).
This visual and logical separation requires each user to explicitly navigate through the levels of abstraction.
3. Establish specific reviews for each level
Project meetings often deal with data and information and knowledge on the same agenda, leading to confusing discussions. A best practice is to organise differentiated reviews:
This differentiation prevents jumping from a debate on the reliability of a figure to a strategic decision without having resolved the first issue.
Performance indicators: what actually changes
Applying this clarification changes the very nature of the indicators tracked. This approach entails:
- Fewer unnecessary indicators;
- More informed decisions;
- Better use of insights.
We no longer debate a figure (“74%, is that a problem?”). We examine the information (gap in the West region) and apply or adjust the knowledge (minimum training required).
Rethinking M&E as a tool for decision-making and learning
The confusion between data, information and knowledge is not merely a matter of terminology. It directly undermines projects’ ability to manage their activities effectively.
In practical terms, this results in:
- An accumulation of data with no real impact on decisions;
- Dashboards difficult to use in steering group meetings;
- A gradual loss of learning from one cycle to the next.
Result: a great deal of effort spent on data collection… for limited decision-making value.
Conversely, the most effective monitoring and evaluation systems are based on a clear structure:
- What is measured (data);
- What is interpreted (information);
- What is capitalised on (knowledge).
Above all, they ensure continuity between these levels and operational decision-making.
It is precisely on this point that Delta Monitoring delivers tangible added value. By structuring monitoring and evaluation flows, the tool enables you to:
- Organise data in a coherent and usable manner;
- Generate information directly useful to steering bodies;
- Capitalise on lessons learnt to avoid repeating mistakes.
RESOURCES
CONTACT
- contact@delta-monitoring.com
- (+1) 514 779 1591