There is a Mi'kmaw concept, Etuaptmumk, that offers one of the clearest guides I know for doing evaluation well in Indigenous health and social services. Articulated by Mi'kmaw Elder Albert Marshall, with Murdena Marshall and the scientist Cheryl Bartlett, it is usually translated as "two-eyed seeing": learning to see with the strengths of Indigenous ways of knowing in one eye, and the strengths of Western ways of knowing in the other, and using both together, for the benefit of all.
What makes it powerful is what it refuses. It does not ask you to choose Indigenous methods over Western ones, which would discard real analytical tools. And it does not do the far more common thing, which is to subordinate Indigenous knowledge to Western method, letting the survey and the statistic sit in judgment over story, ceremony, and lived experience. Two-eyed seeing holds both as legitimate, and asks you to weave them.
Two-eyed seeing is not a compromise between rigour and respect; it is the recognition that the deepest rigour needs both eyes open.
In practice, it changes how an evaluation is built. Indicators are co-designed with the community rather than imported wholesale. Story and relationship are treated as data, not as colour to decorate the "real" findings. Quantitative rigour is welcomed for what it does well, scope, pattern, comparison, without being allowed to crowd out the knowledge it cannot capture. And the community's frame for what success means is kept at the centre, instead of being replaced by the evaluator's.
The default posture of evaluation has long been a one-eyed one, confident that a single way of knowing could see the whole picture. Two-eyed seeing is the discipline of admitting that it cannot, and that the fuller, truer, more rigorous view comes from keeping both eyes open at once. Reference: Bartlett, Marshall and Marshall (2012).