Research

METHODS

We employ rigorous experimental and quasi-experimental designs to generate causal evidence for policy. Below are the four main research methods we use, each linking to representative studies.

Systematic Review and Meta-analysis

A systematic review is a rigorous summary of all studies on a specific question. A meta-analysis is a statistical method, often part of a systematic review, that combines results from multiple studies into one overall estimate. Together, they offer the strongest evidence by improving statistical power and clarifying uncertainty across conflicting studies.

Nature-based interventions & dementia agitation

Nature connection in school curricula

Natural Experiment

A natural experiment uses an external event or policy change to create treatment and control groups by chance. Examples include new metro lines or smoking bans. This allows researchers to estimate causal effects when RCTs are not possible.

Public transit & older adult health

Quasi-experiment

A quasi-experiment is like an RCT but without random assignment. Participants are assigned to groups by non-random means (e.g., self-selection). This makes causal conclusions less certain than RCTs, but quasi-experiments are often more feasible in real-world settings.

Mobile-based intervention for adolescents

Randomized Controlled Trial (RCT)

A Randomized Controlled Trial (RCT) is a scientific experiment design that aims to determine whether a specific intervention (like a new policy, drug, or teaching method) causes a particular outcome. It is considered the gold standard for establishing causality because it eliminates bias.

Mental health & nature connectedness

Parent–child relationship & urban nature

Workplace stress resilience

VR for dementia care