6. The Research Design Ecosystem

From "literature review" to boundary scanning

The phrase "literature review" frames the task as bounded: find what's been published in your area, summarize it, identify a gap. For mono-disciplinary research this works. For boundary-crossing work it doesn't, because the relevant evidence lives in places that don't show up in "your area."

A better frame: boundary scanning — the active practice of looking across disciplinary, methodological, and epistemic boundaries to assemble the evidence base your problem actually needs. The work isn't reviewing; it's hunting.

This shift has practical consequences for how you search, what counts as a source, and how you synthesize.

Searching across disciplinary silos

Disciplinary databases reflect disciplinary boundaries, not problem boundaries. PubMed catches biomedical and a slice of public-health work. PsycINFO catches psychology. ERIC catches education. Sociological Abstracts catches sociology. Web of Science and Scopus catch broad-but-not-everything.

For a transdisciplinary problem, your search strategy should map which database your problem's evidence lives in, then search them all. A reasonable starting set for a public-health-and-social-policy problem might be PubMed + Web of Science + Sociological Abstracts + ERIC + a country-specific government repository.

The corollary: a paper that says "we searched PubMed" without justification is a paper that picked a discipline silently. Justification means naming why the chosen databases cover the problem space.

Gray literature: the evidence that journals don't carry

A surprising amount of the most relevant evidence on applied problems lives outside peer-reviewed journals:

• Government reports — state-level health departments, federal agency analyses, regulatory dockets
• NGO and think-tank papers — RAND, Urban Institute, Mathematica, Brookings, local foundations
• Conference proceedings — often the most current methodological work
• Practice guidelines — clinical society guidelines, professional standards
• Theses and dissertations — frequently the only source for in-depth study of niche populations
• Community-produced reports — community needs assessments, advocacy organization white papers

Excluding gray literature is not rigor; it is a method choice with consequences. The standard transdisciplinary search includes at least one purposive sweep of gray sources, documented for audit.

Honoring methodological diversity in synthesis

The traditional systematic review treats studies as comparable units to pool. This works for narrow biomedical questions and breaks down quickly for transdisciplinary problems, where studies span paradigms, methods, and populations.

Two alternatives:

• Integrative review (Whittemore & Knafl): explicitly admits diverse study designs and synthesizes across them. Useful for emerging fields.
• Critical interpretive synthesis (Dixon-Woods et al.): treats synthesis as an interpretive act, producing a theoretical contribution rather than a summary.

The key shift: synthesis is claim-making, not stacking. The product of a transdisciplinary synthesis is a position about how the evidence fits together, not a table of summaries.

When disciplinary databases will fail your problem

Some clues that a database-driven strategy won't be enough:

• The problem has a strong practice or community dimension that academic literature treats only thinly
• The relevant evidence is recent enough that peer review hasn't caught up (12–24 months for fast-moving fields)
• The phenomenon is named differently across disciplines and your search terms don't capture all the names
• The most relevant work has been done by graduate students whose dissertations weren't published in journals
• The community holding the knowledge has been historically excluded from academic study

When these signals appear, supplement database search with citation chaining, expert outreach, and community partnership. Treat the search itself as a research design step, not a preliminary.

Building the search as a research design move

A transdisciplinary search isn't a literature-review checklist; it's a design decision documented with the rest of the protocol:

1. Problem map — what disciplines, populations, and evidence types are relevant?
2. Database set — which databases are needed, and why?
3. Search terms — including cross-discipline synonyms and idioms-of-distress equivalents
4. Gray-literature plan — which non-academic sources, and how identified?
5. Practitioner/community knowledge plan — what knowledge cannot be found in any database, and how will you elicit it?
6. Synthesis approach — integrative review, critical interpretive synthesis, narrative synthesis, or other

Documenting these decisions makes the search auditable and treats it as part of the research design rather than a preliminary.

A worked example

A research team studying community responses to youth violence is preparing to launch an intervention study. Their boundary-scanning strategy:

• Database set: PubMed (clinical/public health), Sociological Abstracts (community sociology), Criminal Justice Abstracts (criminology), ERIC (school-based interventions)
• Gray literature: City Office of Violence Prevention reports, local foundation evaluations, two community-led needs assessments
• Practitioner knowledge: structured interviews with violence interrupters, school counselors, faith-based program leaders
• Synthesis: critical interpretive synthesis, producing a theoretical claim about which intervention features cross successfully across community contexts

Each of these is documented in their protocol. The published paper that emerges can be audited against this map.

Closing

"Literature review" is the wrong frame for boundary-crossing work. Boundary scanning treats search as a research design step, draws on multiple disciplinary databases, includes gray literature, and admits practitioner and community knowledge. Synthesis is claim-making, not summary-stacking.

Next: literature synthesis as boundary work — appraising across paradigms and writing reviews that take positions.