5.3 Essential IM Techniques
By this point in the Handbook, we’ve explored how IM is structured, how data flows, and how tools are selected. This chapter turns some of those concepts into practice — focusing on the day-to-day techniques that bring IM systems to life. Whether managing registration forms, building databases, or selecting data collection tools, these operational skills are what make IM functional, responsive, and impactful in the field.
These techniques are not only essential for dedicated IM staff, but also for programme managers and MEAL officers who rely on accurate, usable data to steer decisions.
5.3.1 Building and Managing Databases, Registration Forms, and Unique IDs (UIDs)
A well-structured database is essential for organizing programme service delivery, activities, and outcomes over time. The process begins with effective registration forms and the design of a Unique ID (UID) system that prevents duplication and links related data across datasets.
Tips for effective database setup:
Use consistent data structures: Each form should follow standard field types and formats (e.g., text, date, dropdown).
Link related tables: Separate but connected tables (e.g., household and individual-level data) reduce redundancy.
Design forms with analysis in mind: Avoid collecting too much free-text; use codes, categories, and validation rules.
Generate or assign UIDs at the point of registration — ideally auto-generated by the system or using logic based on location, service, and time.
5.3.2 Understanding Unique ID and Database Tech Guidance
A Unique ID is a non-repeating code assigned to an individual, household, collective, facility, or case that allows you to:
Track records across systems
Prevent duplication
Match entries for follow-up, referrals, or reporting
UID options include:
Auto-generated codes from digital platforms (e.g., KoBo, ODK)
Logic-based IDs (e.g., AO-location-year-number)
Database tools vary by context. IMOs should match the tool to the complexity of their needs:
Excel/Google Sheets – small-scale, easy to use, but prone to errors
Access/Airtable – good for relational databases without coding
Kobo/ONA CommCare – mobile-friendly with built-in form design
SQL or cloud-based systems – for larger or multi-office datasets
Always ensure that access is restricted, backups are routine, and that data protection protocols are followed.
5.3.3 Assessments and Data Collection Best Practices
Whether conducting a needs assessment, baseline survey, or ongoing monitoring, strong planning and design are critical.
Key principles:
Start with a clear purpose: What decisions will this data inform?
Use existing data where possible: Avoid duplication — review secondary sources first.
Design for analysis: Ensure questions directly support registration, referrals, and activity planning.
Train data collectors: Ensure ethical conduct, proper consent procedures, and accurate use of digital tools.
Pilot the form: Identify errors in skip logic, question clarity, or survey length.
Plan for indicators and analysis: Identify which indicators the data will support and how it will be analyzed before collecting.
5.3.4 How to Choose the Right Data Collection Tool?
No single tool works for every scenario. Your choice should depend on the following:
Connectivity
Will data collectors have internet access in the field?
Volume
How many respondents or entries are expected?
Data Type
Is the data mostly text, numbers, GPS, media, or sensitive info?
User Skill Level
Are field teams familiar with mobile tools or do they need simpler interfaces?
Data Security
Is encryption or anonymization required?
Integration
Will this tool need to link to another system (Power BI, Excel, database)?
Device Requirements
Does the tool require a laptop, tablet, or smartphone to operate effectively?
Context of Use
Will it be used in the field, in an office, or in low-resource environments?
Common tools include KoBoToolbox, ODK, SurveyCTO, CommCare, and Google Forms — each has trade-offs in usability, cost, and security.
5.3.5 How to Prioritize What Data to Collect?
Collecting too much data can be as harmful as collecting too little. It burdens field teams, frustrates participants, and overwhelms analysis.
To prioritize:
Align every data point with a specific use case — if it won’t be used, don’t collect it.
Categorize data as:
Must-have (critical for reporting or delivery)
Nice-to-have (useful but not essential)
Redundant (already collected elsewhere or never used)
Review past forms to identify unused fields or repeated questions
Limit sensitive data unless absolutely necessary
Follow the principle of data minimization: collect only what you need, when you need it.
5.3.6 How to Clean Data?
Data cleaning is often underestimated — but it’s one of the most impactful IM tasks. Cleaning improves reliability, trust, and usability.
Core cleaning tasks:
Remove duplicates using UID or name+date logic
Flag missing or out-of-range values
Standardize entries (e.g., “F”, “female”, “FEMALE” → “Female”)
Validate against expected values (e.g., number of children shouldn’t be negative)
Use simple tools like Excel filters, data validation, or data cleaning scripts
Cleaning should be planned into workflows — not done as an emergency before a report.
📚 For practical tools, examples, and guidance on cleaning humanitarian data, explore CartONG’s Data Cleaning Module
5.3.7 Key Takeaways
These essential IM techniques form the building blocks of quality data systems.
IMOs should be proactive in designing tools, workflows, and databases that balance simplicity, protection, and usability.
Good data practices are not about volume — they’re about purpose, clarity, and structure.
Every entry collected, stored, and analyzed should serve a decision, insight, or accountability function.
REFERENCES & FURTHER READINGS
Last updated