The four domains
Every practice is assessed across the same four domains at every endorsement level.
Practice Model. The practice's conceptual model and rationale: the need it addresses, the risk and protective factors it targets, the research or theory behind it, and the logic model connecting them, along with its core components.
Practice Implementation. Whether the practice reaches its intended population, how the people delivering it are trained and supported, and how faithfully and how well it is delivered.
Outcomes & Improvement. The meaningful outcomes the practice demonstrates through a sound and appropriate evaluation, and how it uses data to improve.
Practice Sustainability. Whether the practice is built to last: routinized so it does not depend on any one person, and financially resilient.
Theory of change and logic model
Your theory of change is the reasoning behind your practice: the need you're responding to, the risk and protective factors you're trying to affect, and why your activities should move those factors in the right direction. A logic model makes that theory visible, mapping the relationships among a practice's resources, activities, and intended outcomes.
Building a logic model gives staff a shared, concrete picture of the practice, which supports more consistent delivery and points to realistic, near-term measures rather than only the most distant outcomes. Programs build theirs early, with support, and it becomes the foundation for what to measure, what fidelity should look like, and what outcomes to expect.
Implementation fidelity
Fidelity is the degree to which a practice is delivered the way it's designed to be. Without fidelity feedback, programs tend to drift from their model over time without realizing it. Measuring it also helps make sense of outcomes: if results are disappointing but fidelity was strong, the explanation likely lies somewhere other than delivery.
Adherence and competence are two different aspects of fidelity, and both matter. Competence is generally harder to measure than adherence, so we look for practical methods such as participant feedback, peer or supervisor review, or observation where feasible.
Fidelity applies to a practice's core components. It doesn't mean nothing can change: a well-specified practice distinguishes its core from the elements that can be adapted to fit a local context.
Adherence. Whether the core components are actually delivered, as specified and at the intended dose.
Competence. How well they are delivered: the skill, responsiveness, and engagement of delivery.
Pilot outcome data
At Level I a program is not expected to have proven its outcomes. Early pilot data showing promise means results that point in the direction the theory of change predicts, reported with their limits. It signals that a practice is on track to build the outcome evidence Level Ill requires.
Meaningful outcomes
This is the standard at the core of Level III. We're looking for credible evidence that a practice produces real improvement for the people it serves, not a particular p-value or a statistically significant result on its own. A significance test is as much a measure of sample size as of impact. A small practice with an important effect can fail one simply because it doesn't serve enough people, so for community-scale prevention programs it's the wrong bar on its own. Any of these can show meaningful outcomes:
Programs work with the Service to Science team to define what will count as meaningful before looking at results and effects are judged against prevention norms, where a modest effect can matter across a whole population.
Because the standard isn't a single test, credibility rests on how evidence is gathered: well-designed measures, at least a before-and-after comparison, an honest account of who dropped out, attention to other explanations for change, and results reported with their uncertainty.
-
A statistically significant improvement, where the sample is large enough to support one.
-
An effect size agreed in advance to count as meaningful, reported with a confidence interval.
-
An effect size comparable to one established for a similar, evidence-based or established practice in a larger or more rigorous study, for a comparable target population.
-
Reliable, meaningful change for a substantial share of the individual participants served.
-
Consistent effects of meaningful size that repeat across multiple cycles or cohorts.
The last three suit practices with more limited reach.
Sustainability
Sustainability is about whether the practice itself is built to last, not the overall financial health of the organization behind it. A durable practice is built into regular operations, does not depend on any single person to deliver it, and has the funding it needs to continue.
Two things tend to threaten durability: depending on a single champion, so the practice fades when that person leaves, and relying on one grant or funding stream, so it ends abruptly if that source is cut. A durable practice is documented and shared across staff and over time, and is not dependent on any single funding source.
SAMHSA alignment
New Hampshire Service to Science uses SAMHSA's guidance on evidence-based prevention and its Strategic Prevention Framework as reference points. Endorsement is a state-level recognition. It does not place a practice on a national registry.
Learn more about SAMHSA's Strategic Prevention Framework: samhsa.gov/technical-assistance/sptac/framework
NH Service to Science
Facilitated by the Behavioral Health Improvement Institute at Keene State College in partnership with the NH Bureau of Drug and Alcohol Services, Department of Health and Human Services.

