Key Research Terms

Causal inference means deciding whether a program or strategy actually caused a change in outcomes. In CVI, this usually means asking whether a decline in shootings, retaliatory violence, or violent injury happened because of the program itself, rather than because of other factors such as citywide trends, policing changes, seasonality, or unrelated interventions happening at the same time.

The counterfactual is the estimate of what likely would have happened without the program. Because we cannot observe the same person, group, or neighborhood both receiving and not receiving the intervention at the same time, researchers use control groups or comparison groups to estimate that “what would have happened otherwise” condition.

Outcomes are the changes observed during or after a program, such as fewer shootings, stronger school engagement, improved mental health. Program impact refers to the part of those changes that can be attributed to the program itself. Put simply, outcomes describe what changed; impact asks whether the program caused the change. To determine impact, one needs a counterfactual or a causal research design.

A process evaluation examines how a program was implemented (e.g., Was the program delivered as intended?) whereas an impact evaluation examines whether the program itself caused changes in outcomes. Both types of evaluation matter. Process evaluations help explain how a program worked, while impact evaluations help determine whether it made a difference.

A control group is a group of people, places, or communities that does not receive the program and is selected through random assignment. Because chance determines who gets the intervention and who does not, researchers can make a stronger claim that differences were caused by the program.

A comparison group does not receive the program but is not randomly assigned. Instead, it is selected to resemble the program group on important factors such as prior shootings, poverty and demographics. Researchers may use matching or statistical adjustments to make the comparison more credible.

Random assignment means using chance to determine who receives the program and who does not. This helps reduce bias because the groups should be similar before the program begins. Random assignment is often seen as the strongest design for estimating impact, but in CVI it is not always practical or ethical, especially when interventions are designed for entire neighborhoods or social networks.

Internal validity refers to how confidently we can say that the program itself caused the observed results. A study has stronger internal validity when it does a better job of ruling out other explanations for what happened.

A systematic review is a review that uses clear and explicit methods to identify, select, and critically assess all relevant studies on a focused question. Rather than summarizing studies informally, it follows a structured process to synthesize the best available evidence.

Why it matters: High-quality systematic reviews are valuable because they assess the total body of evidence, not just one study.

A meta-analysis is a statistical method that combines findings from multiple studies to estimate the overall size and direction of an effect. When the studies are sufficiently comparable, a meta-analysis can provide a stronger summary than any one study alone.

Why it matters: Meta-analyses often sit near the top of the evidence pyramid (hierarchy) because they pool results quantitatively, but they are only as strong as the studies they include.

A scoping review maps the main concepts, types of evidence, and gaps in research on a topic. It is especially useful when a field is still developing, when definitions vary, or when researchers want to understand what kinds of studies exist before asking narrower questions about effectiveness.

Why it matters: Scoping reviews are helpful for understanding the shape of a field, but they usually do not provide the same level of evidence about effectiveness as a systematic review or meta-analysis.

A summary review is a broader overview of the literature that highlights major findings, lessons, or themes from prior studies. Summary reviews can be very useful for practitioners because they often translate research into more accessible takeaways, but they are much less formal than systematic reviews and meta-analyses and may not use transparent methods for selecting and assessing studies.

Why it matters: Summary reviews can orient readers to a field, but they generally provide a weaker basis for judging effectiveness than systematic reviews or meta-analyses.