King's College London

In the first of our series of blogs from doctoral researchers, George Kinkead argues that securing participation is the make-or-break challenge of quantitative research.

The nuts and bolts of fieldwork often leave PhD candidates caught between a rock and a hard place. Undergraduate research relies on small, convenient samples drawn from other students, while established academics secure funding to properly incentivise participation. Doctoral research, by contrast, frequently depends on sustained engagement from external participants who have little intrinsic stake in the project. How many strong study ideas fail simply because people aren’t interested enough to engage meaningfully? I would argue, then, that securing sufficient participation is the central determinant of quantitative research quality. Here, I discuss some recommendations for success in recruitment; sample size really is the hill researchers are willing to die on.

This hill has the dual terrain of prerequisite power analysis and post hoc significance. We use statistical significance to decide whether an observed effect is likely to be real rather than noise. Statistical power, by contrast, is the probability that we will reach that threshold if a real effect actually exists. In other words, power is our chance of detecting significance when there is something to detect. It is typically calculated based on expected effect size and variability in the data. Small effects require larger samples, much like needing a more powerful microscope to observe smaller microbes. In practice, researchers often aim for around 80% power, meaning an 80% chance of detecting a true effect. If we fall short, the study becomes underpowered: we expend time and effort only to arrive at ambiguity, unable to distinguish between no effect and insufficient data.

The power analysis will tell us roughly how many participants we need; but, like any statistical instrument, it is not rooted in real-world psychology, thus it cannot help explain how do we find this sample. Researchers are far more invested in their projects than participants. What feels like an important, well-designed intervention to us can easily look like just another demand on someone’s time.

At the same time, we have to avoid the pitfalls of coercion or unethical practice. And ideally, we want to observe behaviour in its natural context, which often means making the research feel as unresearch-like as possible. In short, we are aiming for ethical, honest, and engaged participation from people who are not altering their behaviour simply because they are being studied. What follows, then, is a set of principles drawn from my own experience in education-based behavioural research for navigating this balance.

1. Designing Mutual Benefit

In many applied settings, schools, colleges, workplaces, access to participants is mediated by gatekeepers. These stakeholders operate under constraints of time, resources, and competing priorities. Recruitment efforts are therefore far more effective when framed as mutually beneficial.

This can take both material and non-material forms. Financial incentives (e.g. vouchers, institutional payments, or lotteries) can help, but are not always necessary or sufficient. Equally important are non-pecuniary benefits, such as delivering workshops or assemblies, supporting careers events, providing feedback or data insights, or offering resources aligned with institutional goals. To paraphrase JFK’s famous speech: Ask not what your gatekeepers can do for you, ask what you can do for your gatekeepers.

2. Reducing Frictions

Perhaps the preeminent principle in behavioural science is to make behaviour “easy.” In practice, this requires resisting the researcher’s instinct to include ever more measures, checks, and exploratory questions. Multiple scales, repeated measures, and extensive ethical disclosures can make participation burdensome. Researchers should therefore ask: what is essential to identify the effect? What can be removed without undermining the core research question? Reducing length, simplifying language, and minimising cognitive load are practical choices but they are central to maintaining data quality. I’ll be honest, I’m over two years into my PhD, and I still don’t get this one right.

3. Making Participation Attractive

Participation is more likely when individuals perceive clear value. Financial incentives should be used carefully. Excessive incentives risk introducing selection effects (attracting participants for the wrong reasons) or creating undue pressure to participate.

Alternative approaches include emphasising personal relevance (e.g. how the study supports exam preparation), providing immediate, usable benefits (resources, feedback), and framing participation as an opportunity rather than an obligation. Attractiveness, in this sense, is less about maximising rewards and more about aligning the study with participants’ existing motivations. Put yourself in the shoes of a 16-year-old student or an overworked mental health nurse, why would they take part in your research?

4. The Importance of the Messenger

One of the most powerful and often underestimated factors in recruitment is who delivers the message. In my own field experiments, moving from email-based recruitment to in-person engagement increased sign-up rates by 200%. Simple actions, such as visiting a classroom, recording a short video, or being physically present, can transform participation. Seeing the researcher, hearing their intentions, and having the opportunity to ask questions makes the request more credible and harder to ignore.

5. Timing and Attention

Even well-designed studies can fail if they reach participants at the wrong time. Individuals are busy, attention is limited, and willingness to engage fluctuates throughout the day and across contexts. Effective recruitment therefore, requires attention to timing: when are participants most receptive? When are competing demands lowest? When is the decision to participate easiest?

In practice, this involves iteration: sending invitations at different times, tracking responses, and refining strategies accordingly. While this may feel more akin to marketing than research, it reflects a necessary adaptation to real-world constraints. Just ask the charity fundraiser who finally got money out of me on the school run, it just happened to be the day my morning meeting was cancelled.

6. Ethical Boundaries

These strategies raise an important ethical question: where is the line between encouragement and coercion? Researchers must remain attentive to the possibility that incentives become undue inducements, institutional authority creates implicit pressure, or repeated contact becomes intrusive.

Participation must remain voluntary, informed, and free from undue influence. If participants feel compelled to engage, their behaviour may no longer reflect the real-world processes the study aims to capture. There is, therefore, an inherent trade-off. Increasing participation may improve statistical power, but overly aggressive recruitment risks undermining both ethical standards and external validity. The principles outlined here aim to strike that balance, supporting participation without compromising autonomy. We should not seek to coerce participants, but we can do what is in our power to make engagement easier, more relevant, and more worthwhile. And if not, it turns out even Power Rangers occasionally run an underpowered study.