10Min Research Methodology - 15 - How to Propose New and Original Research from Existing Research

TL;DR

Start model-building by selecting recent, hypothesis-driven studies that map clear relationships among variables, then use systematic reviews/meta-analyses to identify research gaps.

Briefing Cornell Notes

Briefing

A practical way to propose genuinely original research is to build a new model from existing studies—then deliberately reduce the risk that someone else will copy the same mediator/moderator combination. The session uses servant leadership as the starting independent variable and shows how to turn “latest research gaps” into a tailored set of dependent variables, mediators, and moderators, while also addressing the common problem of ending up with a model that is too similar to prior work.

The workflow begins with identifying a recent body of research (the example narrows to 2021) and focusing on studies that test relationships among variables—often via survey research and framed through hypotheses. A systematic literature review or meta-analytic review is treated as a gap-finder: these reviews typically highlight where evidence is missing or inconsistent, which becomes the raw material for new research questions and objectives.

From one focal study—“the interplay of green leadership”—the example extracts a concrete structure: green servant leadership is positioned as the independent variable, intrinsic motivation as a key mechanism, and employees’ pro-environmental behavior as the dependent variable. In that study, green self-efficacy functions as a moderator, shaping when or how the leadership–motivation–behavior link holds. The session then demonstrates how to expand the model by adding additional mediators recommended in the limitations/future directions section. Instead of stopping at intrinsic motivation, it adds mediators such as green empowerment, green knowledge sharing, green trust, and green climate—each offered as plausible pathways between green servant leadership and pro-environmental behavior.

Next comes the moderator layer. The example keeps self-efficacy as one option but also considers other moderators suggested by prior research, including green locus of control, altruism, conscientiousness, and green job crafting. The key design principle is theoretical fit: the independent variable must plausibly affect the mediator, and the mediator must plausibly affect the dependent variable.

However, the session flags a major originality problem: even if a model is “new” to the researcher, it may already exist in another paper—especially if other scholars used the same mediators and moderators. The fix is not just to swap variables randomly. Instead, it recommends searching within the broader topic area (e.g., green servant leadership) for additional variables that have not been combined in the same way, or shifting to a different but related research stream.

Concretely, the session suggests using future research directions from multiple papers rather than relying on a single study’s recommendations. One approach is to take mediators from one paper and moderators/outcomes from another, thereby lowering the chance that another researcher will independently assemble the same exact model. Another approach is to introduce comparative leadership styles—such as adding green transformational leadership alongside servant leadership—to test which leadership style better predicts green outcomes. The result is a more complex model that “imports gaps” from several studies, increasing the likelihood of a distinctive contribution while staying grounded in established theory.

Cornell Notes

The session lays out a method for proposing an original research model using existing literature, with servant leadership as the example independent variable. It starts by pulling a recent, hypothesis-driven study that links green servant leadership to intrinsic motivation and pro-environmental behavior, then uses the study’s future directions to add mediators (e.g., green empowerment, green knowledge sharing, green trust, green climate) and moderators (e.g., green self-efficacy, green locus of control). To avoid ending up with a model someone else already used, it emphasizes theoretical justification and then recommends combining elements from multiple papers—such as adding variables from a second study or comparing servant leadership with another leadership style like green transformational leadership. This multi-source gap approach increases originality while remaining defensible.

How does the session turn a literature gap into a usable research model (IV, mediator, DV, moderator)?

It starts with a recent study that tests relationships among variables (often survey-based and hypothesis-driven). From that study, it extracts a workable core: servant leadership (IV) → intrinsic motivation (mediator) → pro-environmental behavior (DV), with green self-efficacy acting as a moderator. Then it expands the model using future research directions: add additional mediators suggested in limitations/future directions (green empowerment, green knowledge sharing, green trust, green climate) and consider other moderators suggested by prior work (green locus of control, altruism, conscientiousness, green job crafting). The model becomes “new” by extending mechanisms and boundary conditions while keeping the causal logic consistent.

Why isn’t changing the data collection technique enough to create a meaningful contribution?

The session warns that swapping the method alone (e.g., changing survey collection technique) doesn’t automatically create a new contribution. A stronger contribution requires new relationships that can be tested. It illustrates this with the idea that replicating the same model in a different country or sector may not add significance unless the relationships themselves change in a theoretically grounded way.

What originality problem can arise even after building a “new” model from one paper’s recommendations?

Even if a researcher assembles a model using mediators and moderators recommended in one study, another scholar may have built the same combination—especially if that paper is widely read. The session frames this as a risk of duplicate models: the same mediators (and possibly the same moderator) could already be in someone else’s future research plan.

What practical strategies reduce the chance of duplicating someone else’s model?

Two strategies are emphasized. First, search within the broader topic (e.g., green servant leadership) for additional variables that can plausibly connect the IV to the mediator and the mediator to the DV. Second, combine gaps from multiple papers: take one mediator or outcome from Paper A and a different mediator/outcome/moderator from Paper B, or add a comparative leadership style (e.g., green transformational leadership) to test which leadership style better predicts green outcomes. This multi-source design lowers the probability that another researcher will independently assemble the same exact configuration.

How should a proposed mediator be justified theoretically?

The mediator must fit a causal chain: the independent variable should plausibly influence the mediator, and the mediator should plausibly influence the dependent variable. The session explicitly instructs checking this logic rather than adding mediators purely for variety.

Review Questions

If you extract mediators and moderators from a single paper, what specific risk does the session identify, and how would you address it?
Using the servant leadership example, propose one mediator and one moderator and explain the causal logic linking IV → mediator → DV and the role of the moderator.
How does combining elements from multiple papers (rather than one) change the likelihood of producing an original research model?

Key Points

1
Start model-building by selecting recent, hypothesis-driven studies that map clear relationships among variables, then use systematic reviews/meta-analyses to identify research gaps.
2
Use the future directions and limitations sections of strong studies to justify adding new mediators and moderators rather than inventing variables without theoretical fit.
3
Ensure every proposed mediator has a defensible causal chain: the IV must affect the mediator, and the mediator must affect the DV.
4
Don’t rely on method changes alone; meaningful contribution usually requires testing new relationships, not just changing data collection technique or location.
5
Treat originality as a design constraint: avoid assembling a model that another researcher could replicate by using the same mediator/moderator set from a single influential paper.
6
Reduce duplication risk by searching within the topic for additional variables and by combining gaps across multiple papers, including comparative leadership styles when appropriate.

Highlights

The example model starts with green servant leadership influencing intrinsic motivation, which then drives employees’ pro-environmental behavior, with green self-efficacy as a moderator.

Future research directions can justify expanding mediators beyond intrinsic motivation to include green empowerment, green knowledge sharing, green trust, and green climate.

A key originality warning: using one paper’s recommended mediators/moderators can still produce a model someone else already planned.

The recommended fix is multi-source gap building—pulling variables from multiple studies and/or comparing servant leadership with other leadership styles like green transformational leadership.