Massey University's Dr John Matthewson, who has trained and worked as a medical doctor and a philosopher, saw an urgent need to get to answer this.
"When I was working as a doctor I was often confronted by an apparent tension," he said.
"On the one hand, policies regarding the best treatment options in a particular setting are based on large population-level studies – but on the other hand, it doesn't matter to any individual patient that this treatment works for ninety-nine percent of people in their situation.
"They care about whether the treatment works for them, and from this perspective, either the treatment is the right one or it isn't."
This potential tension between our very best research and the practical applications of that research had been on his mind ever since.
And Matthewson said the issue wasn't going away: from healthcare to conservation to insurance premiums, policy choices depend on information regarding populations, and big data and population-level statistics would increasingly drive crucial decision-making.
"This has important downstream effects on the individuals involved and society more generally, but it isn't always entirely clear why investigators pick out the groups they do, or whether there is any systematic concept of 'population' underlying research across the various scientific domains," he said.
"For these reasons, I'm particularly interested in answering three questions: How are populations in scientific practice defined and delineated? How are they investigated? How can this type of investigation be used most effectively in further research and application?"
A new study supported with a $300,000 grant from the Marsden Fund would first look at philosophical methods and questions, such as analysing the concept "population" as it was used in various branches of the sciences.
"For example, in a statistician's work a population may just mean 'the group of things we are currently investigating', but 'population' means something more specific than this to an evolutionary biologist, and to an epidemiologist it will mean something different again," he said.
"The real challenge here will be to develop an account that accommodates all of these different uses of the term, while still saying something constructive about what unifies them all."
Next, he'd use this account to carry out a series of in-depth case studies across the natural and social sciences.
"Here I'm interested in exactly why investigators choose the groups they do, and what makes a particular kind of grouping suitable for a particular research task.
"This will involve close reading of published articles, but more importantly I will also need to talk directly to scientists about the decisions they make and the basis of these decisions."
Lastly, he'd consider the practical applications of population-based research.
"Once we have learned from practitioners what makes some grouping appropriate for a specific research question, we can then use this to evaluate how the research is then subsequently employed to inform policy decisions."
Although the philosophy of science was its own independent field of research, he hoped his findings would be just as useful to scientists as anyone else who wanted to draw on them.
"This is why my primary results will be based on discussions with scientists who actually do population-based work, and why the questions raised by the project are not merely about abstract concepts, but also about application," he said.
"So my expectation and hope is that the project will help stimulate and guide further discussion regarding this important topic, in both the academic and public spheres."
