Get AI summaries of any video or article — Sign up free
#SmartPLS4 Series 35 - Moderated Mediation in SmartPLS (See Description) thumbnail

#SmartPLS4 Series 35 - Moderated Mediation in SmartPLS (See Description)

Research With Fawad·
5 min read

Based on Research With Fawad's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

SmartPLS 4’s process tool enables moderated/conditional mediation directly, avoiding the multi-step workaround required in SmartPLS 3.

Briefing

SmartPLS 4 now supports moderated (conditional) mediation directly through a built-in “process” tool, eliminating the workaround that was needed in SmartPLS 3. The core payoff is practical: researchers can test whether an indirect effect changes depending on a moderator—without manually stitching together separate mediation and moderation steps.

The walkthrough uses a simple conditional mediation model with continuous variables: collaborative culture (CC) as the independent variable, assurance (ASR) as the mediator, organizational performance (OP) as the dependent variable, and role ambiguity (RA) as the moderator. The central question is whether RA alters the indirect pathway from CC to OP through ASR. In other words, it’s not only about whether ASR mediates the CC → OP relationship; it’s about whether the strength of that indirect effect depends on the level of RA.

The session frames conditional mediation as a combination of mediation and moderation. Formally, CC influences OP through ASR, while the CC→ASR link is conditioned by RA. The key statistical object is the “index of moderated mediation,” which tests whether the indirect effect is significantly moderated. In this model, SmartPLS 4 provides the index directly, and it is computed as the product of two path coefficients: P2 (mediator → dependent) multiplied by P5 (the interaction term’s effect on the mediator). If this product is significant, the indirect effect is treated as moderated.

Before running the conditional mediation analysis, the workflow still requires standard measurement and structural checks—reliability and validity for constructs, plus structural model assessment—so the conditional mediation results rest on an acceptable measurement model. After setting up the process model in SmartPLS 4 (dragging CC, ASR, OP, and placing RA as the moderator on the relevant path), the analysis proceeds with bootstrapping. The example uses bootstrapping with 500 iterations (noting that 10,000 is commonly recommended), and it reports results using the path coefficients and, most importantly, the specific indirect effects tied to the moderated mediation index.

The results in the example indicate moderated mediation is present: RA significantly moderates the indirect effect from CC to OP through ASR. The session also demonstrates how to interpret the conditional indirect effect at different RA levels. At higher RA, the mediating effect weakens; at lower RA, the indirect effect strengthens. Across mean, high, and low RA levels, the conditional indirect effects remain statistically significant, reinforcing the conclusion that role ambiguity changes the size of the mediated pathway.

Overall, the practical message is clear: SmartPLS 4’s process tool enables conditional mediation testing by focusing on the moderated mediation index (P2×P5) and probing conditional indirect effects across moderator levels—turning a previously cumbersome task into a direct, reportable workflow.

Cornell Notes

Conditional (moderated) mediation tests whether an indirect effect depends on a moderator. In the example model, collaborative culture (CC) affects organizational performance (OP) through assurance (ASR), while role ambiguity (RA) moderates the CC→ASR part of the mediation. SmartPLS 4 provides an index of moderated mediation that, for this setup, is computed as the product of path coefficients P2 (ASR→OP) and P5 (the interaction term’s effect on ASR). A significant P2×P5 indicates the indirect effect is moderated. Probing conditional indirect effects at mean, high, and low RA shows the mediated effect decreases as RA increases and increases as RA decreases, with significance at all levels.

What does “conditional mediation” mean in practice, beyond standard mediation?

Standard mediation asks whether CC influences OP through ASR (i.e., whether the indirect effect via ASR is nonzero). Conditional mediation adds a moderator (RA) that changes the size of that indirect effect. Here, RA conditions the CC→ASR link, so the indirect effect CC→ASR→OP varies with RA levels. The key is not just “is mediation present?” but “does mediation strength change across contexts or individual differences?”

How is the index of moderated mediation computed in the example model?

For this specific setup, SmartPLS 4 uses the product of two paths: P2×P5. P5 is the effect of the interaction term (CC×RA) on the mediator (ASR), and P2 is the effect of the mediator (ASR) on the dependent variable (OP). Multiplying P2 and P5 yields the moderated mediation index; significance of this product indicates the indirect effect is moderated by RA.

Why do reliability/validity and structural checks still matter before conditional mediation?

Conditional mediation relies on mediation and moderation paths that come from the measurement and structural models. The workflow still requires assessing construct reliability and validity (measurement model) and evaluating structural relationships before interpreting moderated mediation results. Without acceptable measurement and structural quality, the conditional indirect effects and the moderated mediation index would be less trustworthy.

What does bootstrapping do here, and what settings were used?

Bootstrapping estimates the sampling distribution of indirect and moderated effects to obtain significance tests. The example runs bootstrapping with 500 iterations (while noting that 10,000 is commonly recommended), uses bias-corrected and accelerated (BCa) bootstrapping, and then inspects the specific indirect effects and the moderated mediation index in the report section.

How should the conditional indirect effect be interpreted at different moderator levels?

After establishing that moderated mediation is significant, the analysis probes conditional indirect effects at RA levels (mean, high, low). In the example, increasing RA reduces the mediated effect (CC→OP through ASR becomes weaker), while decreasing RA increases the mediated effect. The conditional indirect effects are reported as significant at all three RA levels, indicating the moderation changes magnitude rather than eliminating the indirect effect.

Review Questions

  1. In the example, which path coefficient corresponds to the interaction term’s effect on the mediator, and how does it enter the moderated mediation index?
  2. What pattern across low, mean, and high RA levels would you expect if role ambiguity weakens the indirect effect?
  3. Why is it not enough to test only the mediation effect when the goal is conditional mediation?

Key Points

  1. 1

    SmartPLS 4’s process tool enables moderated/conditional mediation directly, avoiding the multi-step workaround required in SmartPLS 3.

  2. 2

    Conditional mediation tests whether an indirect effect (X→M→Y) changes as a moderator (W/Z) changes.

  3. 3

    In the example, collaborative culture (CC) affects organizational performance (OP) through assurance (ASR), with role ambiguity (RA) moderating the CC→ASR link.

  4. 4

    The moderated mediation index is assessed using the product of path coefficients P2×P5, where P5 is the interaction’s effect on the mediator and P2 is the mediator’s effect on the dependent variable.

  5. 5

    A significant P2×P5 indicates the indirect effect is moderated, meaning the mediation strength depends on RA.

  6. 6

    Conditional indirect effects should be probed at mean, high, and low moderator levels to show how the indirect effect changes across contexts.

  7. 7

    Measurement-model reliability/validity and structural-model assessment should be completed before interpreting conditional mediation results.

Highlights

SmartPLS 4 computes the moderated mediation index directly in the process workflow, using the product of key path coefficients (P2×P5).
Role ambiguity (RA) weakens the indirect effect from collaborative culture (CC) to organizational performance (OP) through assurance (ASR) as RA increases.
Probing conditional indirect effects at mean, high, and low RA levels shows the mediated pathway strengthens when RA is lower and weakens when RA is higher.
The analysis relies on bootstrapping (BCa) to test significance of moderated mediation and conditional indirect effects.

Topics