Data Collection (Survey)
Responsibility
Responsibility for data collection, costs and timing must be clearly outlined in the project’s M&E plan. Generally, indicator data collection, including survey set-up and initial analysis, is completed by the project, organised by the M&E person with support from the project team or by the implementing. The Country Office should ideally lead the result aggregation and consolidation across projects for the performance indicators, ensuring that data is consistently reported and visualised or used on a country programme level.
When planning a new country strategy respectively the way how to measure it make sure to make use of the performance indicators. It will make your live easier
Survey as the Data Collection Method
For most outcome indicators, we use surveys (household surveys, KAP surveys, agricultural surveys, perception survey, etc.). Sometimes the indicator is a measure that compares outcome with baseline values, in that case a longitudinal survey must be implemented, this means that the same persons or households get interviewed at different points in time. In most cases however, the survey is based on a cross-sectional sample, meaning that for each survey (baseline, annual outcome monitoring, midline, and/or endline), a new sample is sampled.
Sampling: When collecting data from all project participants is not feasible, a representative sample can be selected to extrapolate findings to the larger target population. Sampling ensures cost-effective and reliable data collection while maintaining accuracy. The sample size must be large enough to provide statistically valid results, typically at a 95% confidence level with a 5% margin of error. The required size depends on factors such as population size, variability in responses, and the level of precision needed. Sample size calculators can help determine the appropriate size. There are different methods to ensure representative samples in survey data collection, the most common being: random sampling, stratified sampling, and cluster sampling.
Random sampling ensures that every individual in the population has an equal chance of selection, minimizing bias and increasing representativeness.
Stratified sampling divides the population into subgroups (e.g., gender, age, geography) and selects samples from each group, ensuring diversity and proportional representation.
Cluster sampling selects entire groups (e.g., schools, villages) rather than individuals, making it particularly useful for large or geographically dispersed populations.
Systematic sampling selects every nth individual from a list or sequence, providing a simple and efficient alternative to pure random sampling.
Convenience sampling involves selecting individuals who are easiest to reach, though it risks bias and may not fully represent the target population. However, in some situations e.g. flooding in some areas and population cannot be reached.
When sampling, ensure that the sample size is appropriate for the population you are studying. For example, suppose you are running activities for two different groups: a) Small-scale farmers who are being supported to increase both their production and income while adopting environmentally friendly farming practices and b) Participants in income-generating activities (IGA), such as beekeeping, where the focus is solely on increasing income. Even though both programs have income-related indicators, the two target groups may be different as they were participating in different project activities. If the groups do not overlap, you must take separate representative samples for each to accurately measure their specific indicators. This is because each group represents a distinct statistical population.
Challenges and Considerations for Surveys as a Data Collection Method: Surveys are a powerful tool for collecting data, but they come with challenges that projects often face. Limited or non-existent budgets, lack of capacity (in terms of time or skills), and operating in difficult contexts, such as fragile or authoritarian environments, can hinder effective survey implementation. To address these challenges, careful planning and budgeting are essential from the outset. Allocating resources early ensures that surveys are feasible.
Cost savings can often be achieved by doing the survey planning and analysis in-house, while hiring local enumerators, local stakeholders or volunteers for data collection. This approach not only reduces expenses but also enhances organizational learning and builds local capacity. When in-house capacity is insufficient, engaging HO or external consultants may be necessary to ensure high-quality data collection. Leveraging partnerships with other NGOs or academic institutions can also be a valuable strategy, allowing projects to co-develop survey methodologies and share resources effectively.
To learn more about different data collection methods, refer to the Pamoja page on Tools and Methods