Reading Between the Lines
Strategies to help address social desirability bias in research
The challenge with researching sensitive topics like gender inclusion is that it’s hard to get honest answers. Under the weight of social and moral expectations, it’s difficult for people to share what they really think. Instead, they end up saying what they believe the “right” answer is.
In research, this is known as social desirability bias and occurs when people, either consciously or unconsciously, answer in a way they think is “acceptable” rather than reflecting their true beliefs.
When we ask about gender in particular, it’s not just a simple question—it’s almost like a moral test. It’s as if we’re asking, “Are you a good person who believes in equality?” The answer is almost too easy to guess. And that’s a problem.
Why does this matter?
As researchers, we aim to uncover real opportunities for change by understanding what people truly think and feel. But when responses are shaped by social desirability, it leaves us in a tough spot. We walk away having gained little meaningful insight into the audience we aim to serve, often leading us down one of two paths: thinking there that is no real issue or doubting the reliability of our data. Take gender bias—we know it’s deeply rooted in society, but if all we get are “morally right” answers, we’re left to guess what the audience truly thinks, shaping the narrative ourselves. We end up relying on assumptions, designing based on what we think we know, rather than the true, nuanced experiences of those we aim to help. The result? Well-intentioned interventions that sound good but miss the mark, too broad to truly connect with anyone.
It’s true that not everyone will fall into this trap. Some individuals will be confident enough to navigate complex dynamics and express their true thoughts; even extreme ones, like dismissing the need for girls' education. But these extreme views are, by nature, challenging to design interventions for. The real opportunity lies with the majority: those who are uncertain or under-confident, simply going along with the status quo. These are the individuals who stand to benefit most from awareness and behaviour change programs, yet, they are also the ones most likely to struggle with answering honestly. Without understanding their true thoughts, we can't design interventions that truly resonate and inspire lasting change.
When we first jumped into researching gender inclusion at AWCs, we knew we had to find ways to overcome some of these biases. We’ve been experimenting with different methods, learning from both our successes and our missteps. Here’s a look at a few methods that have worked well for us so far:
Observation Study
Observation is a core method in ethnographic research that allows researchers to gather data by immersing themselves in the natural environment of the group or community they are studying. It is particularly valuable for understanding social practices, behaviours, and interactions in real-time and in context.
Our first thought was straightforward; we decided that instead of relying on what people say they do, we’ll observe what they actually do. So, we went into AWCs with an observation tool in hand, ready to dive into the spaces, resources, AWW communication, children’s participation, and TLM—all through a gender lens. The tool was designed to guide researchers during their observation, while giving them the freedom to capture notes in their own way. Each day, our team immersed themselves in the AWC, arriving before the children and staying long after they left, observing every subtle interaction with keen attention.
Despite our best efforts to blend in, we stood out. Given our presence, understandably, most AWWs were on their best behaviour. We made sure to communicate that we weren’t there to supervise or judge them but to engage with the children. Researchers who were able to immerse themselves naturally—walking around and playing with the kids—found it easier to put the AWWs at ease, helping them be more authentic. We soon realised that sitting in the corner, notebook in hand, wasn’t the best approach. The observation tool sometimes worked better as a guide—something to review beforehand and fill out afterward, so the AWWs didn’t feel like they were being graded.
This approach was incredibly time-intensive. We spent entire days at the AWCs, watching and absorbing every interaction that unfolded. Eventually, it became clear that it would be impossible for just my teammate and me to visit many centers across such diverse regions. We were able to bring in team members from all over the country and add diversity, without which our observations would have been limited to only the geography and social context we had access to.
With multiple researchers comes the challenge of personal bias. Each person carries their own lens, which can complicate efforts to capture an authentic picture. To address this, we made sure to recruit people we trusted—those who understood the mission deeply. We also ran a comprehensive sensitisation workshop to align everyone with the goal. The observation tool, with its specific prompts and questions, helped a lot in keeping the focus clear. Some researchers found it limiting, occasionally writing outside the lines, but for others, it provided essential structure. We kept our notes strictly observational—recording only what we saw and heard, saving thoughts and interpretations for the debriefs afterward.
The debriefs were where the real insights emerged. Discussing our experiences while they were fresh in our minds allowed us to spot patterns that might have otherwise slipped by. For example, nearly all of us noticed the presence of girls over the age of 6 at the AWCs; an observation that led to important insight into under investment in girls education.
Card Sorting
Along with observation, we needed to speak with AWWs and families, but we knew simply asking direct questions wouldn’t lead us to the right answers. We borrowed from a UX design research technique and created a deck of cards featuring gendered statements like "girls are better suited for leadership roles than boys."
Card sorting is a UX design method used to understand how users categorise and organise information. It involves giving participants a set of cards, each representing a piece of content, and asking them to group or label these cards in a way that makes sense to them.
We asked participants to sort these cards into three categories: agree, disagree, or don’t know; while doing so we engaged in casual conversation across several topics including investment in education, their opinions on career aspirations, sharing household work among others. Here are a few key takeaways from the process:
Beyond just providing data through observing where cards were placed, this activity helped break barriers and led to deeper, more complex conversations. AWWs shared stories about parents who chose better education for their sons or proudly stated that their sons didn’t participate in household chores. These conversations serve as windows into the communities AWWs work with, providing insights that would be difficult for us to uncover on our own.
AWWs, workers, and family members typically engaged in these discussions in groups, which allowed for open conversations. Occasionally, disagreements arose, and as one tried to convince the other, we caught glimpses into their thought processes. In one instance, while sitting with a young mother in Dhule, Maharashtra, we chatted as she organized cards in front of her. Though a 10th-grade graduate herself, she expressed the importance of giving girls and boys equal opportunities. Our conversation was open and candid, until her mother-in-law came home. Suddenly, she lowered her voice and stopped sharing her opinions. Despite believing in gender equality, she confided that her daughters wouldn’t have the chance for higher education, as they would be married off early. The insight for us: To spark real change, educating young parents about gender inclusion isn’t enough. Our interventions must target the family’s decision-makers, often the grandparents of the children in our programs. We need to acknowledge that providing equal opportunities for sons and daughters often faces societal pressures and household conflicts. Parents need to be empowered to have these conversations at home, and a young mother’s own confidence is key to that.
Studying body language played a crucial role as well. AWWs sometimes placed biased cards firmly in the "disagree" column, passionately explaining their reasoning to the helpers. This subtle act highlighted the potential of AWWs as change-makers and revealed those who could become true champions of a gender inclusion program. With the right training, these AWWs could become powerful advocates for change.
Across every method we’ve explored, one profound truth emerges: the real insight is found not just in the data itself, but in the art of discerning what lies beneath the surface. It’s in the mindful, intentional journey of applying these tools—where subtle nuances come alive—that we unlock deeper understanding. To truly grasp complex topics, we must embrace the delicate balance between method and meaning, reading between the lines to see what often goes unnoticed.