03 · What You Need to Know
Integration Is Not the Same as Making Disciplines Say the Same Thing
Begin with a shared problem rather than a demand for consensus
Interdisciplinary research usually begins because a problem cannot be understood adequately from one disciplinary perspective alone. That does not imply that the relevant disciplines already agree about the problem.
They may define it differently. They may prioritize different outcomes. They may explain it at different levels. They may disagree about appropriate methods or evidence.
The starting point for integration should therefore be a sufficiently shared problem, not an assumption that all disciplinary accounts must converge on one interpretation.
Research on interdisciplinary integration commonly emphasizes establishing common ground among disciplinary insights. Common ground is useful because it allows perspectives to become mutually intelligible and connected. It should not be confused with complete agreement.
First determine what the disagreement is actually about
Researchers sometimes describe disciplinary disagreement too broadly: “psychology sees it differently from sociology,” or “the quantitative and qualitative perspectives conflict.” Those statements tell you very little.
Locate the disagreement precisely.
| Possible source of disagreement |
Question to ask |
| Definition |
Are the disciplines using the same term for different constructs? |
| Object of study |
Are they actually explaining the same phenomenon? |
| Mechanism |
Do they propose different explanations for the same outcome? |
| Level of analysis |
Is one explaining individuals while another explains groups, institutions, or systems? |
| Boundary conditions |
Could both accounts be valid under different circumstances? |
| Evidence |
Do they disagree about what observations can support the claim? |
| Research aim |
Is one trying to predict while another seeks interpretation, explanation, design, or critique? |
Once the disagreement is specified, integration becomes a tractable intellectual task rather than an abstract request that researchers “combine perspectives.”
Do not erase differences by standardizing terminology too early
A common strategy in interdisciplinary teams is to agree on one vocabulary. This can help communication, but premature standardization can conceal conceptual differences.
Suppose two fields use “engagement.” One treats it primarily as an individual psychological construct. Another conceptualizes engagement as relational and situated within learning activity. Calling both simply “engagement” does not create common ground if the underlying meanings remain different.
Begin by preserving the definitions long enough to compare them.
Terminological agreement
Researchers use the same word.
Conceptual common ground
Researchers understand where their concepts overlap, differ, and can be related sufficiently for the research problem.
The second is considerably more useful than the first.
Look for relationships other than equivalence
Researchers often approach integration as though two disciplinary insights must either mean the same thing or one must be wrong.
There are many other possibilities.
One concept may be broader than another. Two constructs may partially overlap. One explanation may concern a mechanism while another identifies the conditions under which that mechanism operates. One discipline may explain individual variation while another explains why that variation is patterned institutionally.
Integration can therefore involve specifying relationships among insights rather than collapsing them.
| Relationship |
Possible integrative response |
| Equivalent |
Translate terminology while preserving disciplinary provenance where useful |
| Overlapping |
Identify shared and distinct dimensions rather than merging the concepts completely |
| Complementary |
Connect the insights as explanations of different necessary parts of the problem |
| Different levels |
Specify a cross-level relationship |
| Competing |
Identify predictions, evidence, or conditions that distinguish the explanations |
| Conditional |
Specify when or where each account may apply |
| Difficult to reconcile |
Preserve the disagreement explicitly and examine its consequences for the research question |
This approach avoids treating synthesis as conceptual averaging.
Common ground can be narrower than either disciplinary perspective
Sometimes the disciplines cannot agree on a broad construct but can agree on a more limited phenomenon.
For example, researchers may disagree about the theoretical nature of “trust” but agree that a particular observable behavior involves a person choosing to rely on an automated recommendation under specified conditions.
That narrower point of agreement may be enough to design a study investigating how each disciplinary account explains the behavior.
Common ground therefore does not need to resolve every philosophical disagreement. It needs to support the particular integration required by the research question.
Preserve different levels of explanation when they are genuinely different
Consider a study of students' use of generative AI.
A psychological perspective may explain individual differences in reliance through cognitive or motivational processes. An institutional perspective may explain how assessment rules create incentives for particular forms of AI use.
Forcing these into one level would weaken both. Institutional policy is not simply another student-level variable, and student cognition is not reducible to institutional structure.
A stronger integration specifies the relationship: institutional conditions may shape the situations in which particular cognitive processes become consequential.
The perspectives remain different. The integration lies in the cross-level explanation.
Allow evidence to remain heterogeneous when the questions require it
Interdisciplinary projects can involve different forms of evidence. The temptation is to impose one evidentiary standard across the entire project so that everything appears methodologically unified.
That can be inappropriate.
A computational analysis may establish properties of a technological system. Interviews may investigate how participants interpret those properties. An experiment may test whether manipulating a property changes behavior. Institutional documents may establish policy conditions.
These forms of evidence support different claims. Integration occurs when those claims are related to the overarching problem, not when every form of evidence is converted into the same methodological language.
Integration can involve productive disagreement
Recent philosophical work on interdisciplinary research argues that differences among disciplinary epistemic systems can produce conflicts concerning aims, methods, evidence, and other epistemic values. Such conflict is not necessarily evidence that interdisciplinary work has failed.
A disagreement can expose assumptions that would otherwise remain invisible.
One discipline may challenge another's definition. Another may reveal that an apparently universal mechanism depends on context. A third may show that a technically successful solution produces a poor outcome when judged according to another criterion.
If the project preserves those challenges and uses them to refine the explanation, disagreement becomes part of integration rather than an obstacle to it.
Watch Out
Do not resolve interdisciplinary disagreement merely by allowing the most institutionally powerful, methodologically dominant, or familiar discipline to define the common vocabulary and acceptable evidence. Apparent consensus can sometimes reflect disciplinary capture rather than successful integration.
Integration requires enough understanding of the other perspective
Researchers cannot integrate a disciplinary perspective they understand only as a handful of borrowed variables.
Work on interdisciplinary collaboration has emphasized that disciplinary perspectives guide how experts use and produce knowledge, while those perspectives can remain largely implicit to the experts themselves. Making them explicit can improve mutual understanding and support the construction of integrated conceptual models.
You do not need to become a second disciplinary specialist overnight. You do need enough understanding to know what a concept means in its original intellectual context, what assumptions accompany it, and what claims it can legitimately support.
If the project requires expertise beyond what the existing researchers possess, this strengthens the case for involving collaborators from the relevant disciplines while the design remains changeable.
Sometimes disagreement reveals internal inconsistency
Not every conflict can be handled by saying that both perspectives are useful.
Two frameworks may make genuinely incompatible assumptions about the central phenomenon. One may require a construct to be stable while another treats it as constituted dynamically by context. One may assume a causal direction that another explicitly reverses.
At that point, determine whether the disciplinary assumptions make the research question internally inconsistent.
The response might be to revise the question, select one conceptualization for a specific purpose, separate the constructs, test competing explanations, or preserve the conflict as part of the research problem. Simply declaring both perspectives complementary is not enough.
The integrated result should be more than a compromise
Compromise and integration are not identical.
If Discipline A claims X and Discipline B claims Y, choosing something halfway between X and Y may satisfy the meeting but produce no defensible explanation.
Interdisciplinary integration should instead clarify what each insight contributes and produce an advancement in understanding that depends on their relationship. Earlier scholarship describes integration in terms of substantive linkages among component analyses, while more recent work emphasizes interweaving perspectives and interconnecting knowledge throughout the research process.
The outcome may be a revised explanation, a cross-level model, a set of boundary conditions, a differentiated concept, a new question, or even a clearer account of why the disagreement cannot currently be resolved.
Intellectual honesty is sometimes a better interdisciplinary product than synthetic harmony.
04 · A Practical Example
Integrating Two Different Accounts of Student Reliance on AI
Hypothetical Example
Why do students continue relying on inaccurate AI-generated feedback?
A research team approaches the problem from psychological and educational perspectives. The psychological account emphasizes students' trust and metacognitive judgments. The educational account emphasizes feedback practices, task design, and students' opportunities to evaluate feedback.
The team initially tries to create agreement by describing all of these as “factors affecting AI trust.” That simplifies the framework, but it also distorts the educational perspective by converting contextual learning conditions into individual trust variables.
A better integration preserves the difference.
Clarify each account The psychological perspective explains how students judge the credibility of AI output. The educational perspective explains how learning and assessment conditions create opportunities and incentives for evaluating that output.
Identify common ground Both perspectives concern whether students critically evaluate AI-generated feedback before acting on it.
Preserve the difference Trust remains a learner-level construct, while assessment conditions remain contextual rather than being relabeled as psychological variables.
Specify the relationship The integrated question asks whether assessment conditions alter when and how students' trust judgments translate into reliance on inaccurate feedback.
Produce an integrated explanation The final account can explain both variation in students' judgments and why the consequences of those judgments differ across educational conditions.
The disciplines did not agree that the problem was fundamentally psychological or fundamentally educational. They established enough common ground to investigate a relationship between the two explanations.
07 · A Quick Checklist
Before Claiming That Disciplinary Perspectives Have Been Integrated, Check This
Before finalizing the integration, check:
State the shared research problem without assuming that the disciplines define every part of it identically.
Compare the definitions, mechanisms, levels of analysis, assumptions, and evidentiary standards underlying each perspective.
Identify precisely where the perspectives agree and where they genuinely differ.
Establish only as much common ground as the research question actually requires.
Preserve concepts or forms of evidence that remain meaningfully distinct rather than forcing them into one vocabulary.
Specify whether disciplinary insights are equivalent, overlapping, complementary, cross-level, competing, conditional, or difficult to reconcile.
Check whether one discipline has implicitly defined the common ground entirely on its own terms.
Show what new explanation, distinction, boundary condition, question, or understanding emerges from relating the perspectives.