M&E / MEAL

What interviewers look for

• You understand the results chain and can place any activity, output or outcome of the advertised project at the right level.
• You can write SMART indicators with a baseline, a target, a data source and a frequency.
• You have really used a mobile data tool such as KoboToolbox or ODK, and you can clean and summarise data in Excel.
• You check data quality and report bad results honestly instead of adjusting them.
• You know how a complaint and feedback mechanism works, including how sensitive complaints are handled confidentially.
• You turn data into short, clear findings and recommendations that the programme team can actually use.

Questions they ask

1. 1“Explain the results chain for this project.”

   Why they ask: Whether you can link what the project does to the change it wants, and whether you understand that outcomes depend on assumptions outside the project's control.

   How to answer

   • Name the levels in order: inputs, activities, outputs, outcomes, impact.
   • Use the project in the advert for each level, not a textbook example.
   • Mention one assumption, such as prices staying stable or people being able to reach markets.

   Take a cash-for-nutrition project. The inputs are funds, staff and the mobile money partner. The activities are registering families with children under five and sending monthly transfers. The outputs are the transfers delivered and mothers trained on child feeding. The outcome we want is better diets for young children, measured by how many food groups they eat. The impact is less malnutrition in the district. Each step depends on assumptions, for example that food is available in local markets at stable prices. If prices jump, the outputs can be delivered and the outcome still missed.

2. 2“How would you design the baseline survey for this project?”

   Why they ask: Whether you can plan a survey from start to finish: indicators, sample, tool, enumerators, consent and checks, and not only fill in forms.

   How to answer

   • Start from the indicators: every question in the tool must feed an indicator.
   • Choose the sample: the population, the sample size, and how households are picked at random in each district.
   • Build the form in KoboToolbox with Somali wording, skip logic and constraints, then pilot it.
   • Train enumerators on consent and neutral questions, and check the data every evening.

   I would start from the logframe and list the indicators the baseline must measure, so the tool has no extra questions. With my manager I would agree the sample: households in the target villages, chosen at random from the registration list in each district. I would build the form in KoboToolbox with the questions in Somali, skip logic and limits on numbers, and test it with a few households outside the sample. Enumerators would get a day of training on consent and neutral questions. During collection I would check each day's data for duration, GPS and missing answers, and call the team leader about any problem.

3. 3“An enumerator submitted many surveys in one afternoon, and the answers look almost the same. What do you do?”

   Why they ask: Whether you know how to detect poor or invented data, and whether you handle it fairly and firmly without accusing anyone in public.

   How to answer

   • Check the metadata: start and end times, interview length, GPS points.
   • Call back a sample of the households to confirm the interviews happened.
   • Talk privately with the supervisor and the enumerator, and hear their side.
   • Remove invalid records, redo those interviews, document the case and inform your manager.

   In a nutrition survey I noticed one enumerator had many interviews that each lasted only a few minutes, with GPS points very close together. I did not accuse him in front of the team. I called back a sample of his households, and several said nobody had visited them. I shared this privately with the team leader and the enumerator, who admitted he had filled forms at a tea shop because of the heat. We removed all his records for that day, sent another enumerator to redo them, and I reported the case to my manager, who handled the disciplinary side.

4. 4“How does post-distribution monitoring work?”

   Why they ask: PDM is one of the most common M&E tasks in cash and relief projects. The panel checks timing, sampling, the key questions and who makes the calls or visits.

   How to answer

   • It is done some days or weeks after a distribution, with a random sample of recipients.
   • Key questions: did you receive the full amount, did anyone ask you for money, how did you use it, did you feel safe, do you know how to complain.
   • It is collected by M&E staff, not by the team that distributed, so people speak freely.
   • Findings go back to the programme with actions, and fraud or abuse is referred at once.

5. 5“How should a complaint and feedback mechanism work, and what do you do with a sensitive complaint?”

   Why they ask: MEAL staff often run the hotline and the complaints log. The panel checks that you know the full cycle, and that sensitive cases never sit in a normal spreadsheet.

   How to answer

   • Offer several channels: hotline, help desk, suggestion box and visits, with a female staff member women can reach.
   • Log, categorise, send to the right person, respond within a set time, and tell the person the result.
   • Sensitive complaints, such as sexual exploitation or fraud by staff, go the same day to the PSEA or safeguarding focal point, with restricted access.
   • Report trends every month so the programme can fix problems that keep coming back.

   Our hotline had a simple log in Kobo: date, channel, location, category and status. Normal feedback, like a missing name or a late transfer, went to the project officer and had to be answered within a week; we then called the person back and closed the case. Sensitive complaints were different. When a caller said a staff member had asked for part of her cash, I took only the basic facts, did not put her name in the shared log, and sent it the same day to the safeguarding focal point. Each month I shared a summary of complaint types with the programme team.

6. 6“Which tools have you used for data collection and analysis, and at what level?”

   Why they ask: The panel will test the tools you name. They want honest levels and a real example, not a long list.

   How to answer

   • Name each tool and what you did with it: built forms, collected data, cleaned data, made charts.
   • Give your real level, for example “confident with Excel pivot tables, basic in Power BI”.
   • Say what you are learning now and how you are learning it.

   I have used KoboToolbox the most. I built forms with skip logic and constraints for a household survey and exported the data to Excel. In Excel I am confident with cleaning, COUNTIFS and pivot tables, and I make the charts for our monthly report. I have used ODK Collect on the phone as an enumerator, but I have not built ODK forms myself. I am at a basic level in Power BI; I am taking a free online course and have built one practice dashboard from our old survey data.

7. 7“Your manager asks you to report a target as met because the donor visits next week. The data shows it is not. What do you do?”

   Why they ask: Integrity under pressure is the core of MEAL. The panel checks that you refuse politely, offer a better option and know how to escalate.

   How to answer

   • Explain calmly that you can only report what the data shows.
   • Offer the true figure with the reason and a recovery plan, which donors accept far better than surprises.
   • If the pressure continues, raise it through the proper channel, such as the MEAL manager or the whistleblowing line.

Example answers are in English, the language most panels use. Say it in your own words.

Topics to revise

• Results chainThe logic from inputs to activities, outputs, outcomes and impact. Outputs are what the project delivers; outcomes are the change in people's lives or behaviour. Expect to place examples at the right level.
• LogframeA table that sets out the goal, outcomes, outputs and activities, with indicators, means of verification and assumptions for each. You may be asked to fill in or fix a row.
• SMART indicatorsSpecific, measurable, achievable, relevant and time-bound. “Improved hygiene” is not an indicator; “% of households with soap and water at the handwashing place, from baseline to target by month 12” is.
• Indicator reference sheetA short page for each indicator: definition, how it is calculated, data source, frequency, disaggregation and who collects it. It stops two staff members from measuring the same indicator in different ways.
• Baseline, midline and endlineMeasurements before the project starts, in the middle and at the end, using the same indicators and similar methods so they can be compared. Without a baseline you cannot show change.
• Sampling (random, stratified, cluster)Random sampling gives every household an equal chance; stratified sampling draws separately from groups such as districts; cluster sampling picks villages first, then households. For a large population, 95% confidence and a 5% margin of error need about 384 respondents, plus extra for refusals.
• KoboToolbox and ODKFree tools for collecting data on phones, offline too. Know skip logic (a question appears only when it applies), constraints (limits on answers), required questions, GPS, and how to export to Excel.
• Data quality checksChecks for accuracy, completeness, consistency and timeliness: interview length, GPS, duplicates, impossible values, spot checks and call-backs. Be ready to name the checks you have run.
• Post-distribution monitoring (PDM)A follow-up with a sample of people after a distribution or transfer: did they receive the full amount, how did they use it, were they safe, did anyone ask them for money.
• Complaint and feedback mechanism (CFM)The channels and steps for receiving, logging, answering and closing complaints and feedback. Sensitive cases go straight to the safeguarding focal point and never into the open log.
• DisaggregationSplitting results by sex, age, disability and location so you can see who is left out. Many agencies ask about disability with the Washington Group questions.
• FGD and KIIFocus group discussions (six to ten people, with a facilitator and a note-taker) and key informant interviews (one person who knows the subject well). They explain the “why” behind survey numbers.

Practical tasks you may get

1. 1A written test: a project summary and 60 minutes to draft a small logframe, or three indicators with baseline, target, source and frequency. Practise on two or three real project descriptions from vacancy notices.
2. 2An Excel test: a messy dataset to clean (duplicates, blanks, wrong codes), then a pivot table by district and sex, one chart and three sentences on what it shows. Practise on any public dataset or one you create.
3. 3Build or fix a short KoboToolbox form, sometimes live on a laptop: a consent question, skip logic and a constraint. Open a free account and build a ten-question household form before the interview.
4. 4A case study: design a PDM or a small survey for a cash project, covering the sample, the questions, the timing, who collects the data and how the findings are used.

Mistakes to avoid

• Calling activities results, or mixing up outputs and outcomes.
• Listing tools you have only watched videos about. Panels test them; give your honest level.
• Saying you would delete outliers or “adjust” data to match the target.
• Forgetting consent and data protection, such as sharing files with names and phone numbers on WhatsApp.
• Reports full of tables that never say what the numbers mean or what should change.
• Not splitting results by sex, age and disability, so nobody sees who was missed.

Quick check

5 questions. Answer each one to see the explanation.