Cluster sampling.
Instrumental variables.
Composite indices.
These were not abstract terms to me—they were the language of my economics education. I spent years working through models designed to answer urgent development questions whether it was how social insurance stabilizes livelihoods or how tax redistribution can uplift marginalized communities. Monitoring and evaluation, as I learned it, relied on structure and clarity. Ask the right questions, minimize bias, and the data will speak.
Conceptually, this approach made sense. On paper, direct questioning seemed more than sufficient. But my experience conducting fieldwork as an evaluator has fundamentally reshaped how I understand data collection, survey design, and the human dimensions of development work.
While working with Kriti Social Initiatives, an organization supporting women’s self-help microfinance groups, I am assisting with monitoring and evaluation during the screening phase, a critical step in assessing financial suitability and understanding members’ household financial contexts.
To carry out this process, Kriti relied on surveys that included sensitive questions regarding monthly income, rent, and household expenses. In theory, these questions should have provided the full story on their socioeconomic status. In practice, they often produced resistance, confusion, or unreliable responses. Some women stiffened as soon as numbers were mentioned. Others laughed nervously, eyes drifting toward husbands or mothers-in-law for reassurance. Many simply did not know how to answer. For women who had never managed household finances directly, income-based questions were not just intrusive—they were genuinely confusing.
It quickly became clear that the direct approach I had been trained to rely on was not working. Through repeated visits and trial and error, our team learned that asking around a question was far more revealing than asking it outright.
One of the most effective changes we made was introducing asset-based indicators. Instead of asking how much a household earned, we asked what they owned. Did they have a cooler? An air conditioner? Both? These seemingly casual questions opened easier conversations and offered clearer signals. A cooler, common and affordable, suggested one income bracket; an AC unit, costly and less prevalent, pointed to another.
We applied the same logic to vehicle ownership. When a respondent reported owning an auto-rickshaw or car, we followed up by asking whether it was owned outright or leased. Those leasing vehicles, often through platforms like Uber, tended to have more stable or higher-paying income sources. These indirect questions consistently yielded more accurate insights than direct financial inquiries ever had.
What surprised me was how much my own mindset began to shift during this process. I started thinking less like a student solving a problem and more like an observer piecing together a story—listening for what went unsaid, noticing surroundings, and paying attention to hesitation, ease, and context.
While I always understood that classroom theory has its limits, fieldwork revealed just how incomplete technical foundations can be without lived context. Being on the ground is what separates a superficial survey from a deep, nuanced, and truly comprehensive one.
