How to Report Moderation Analysis Results from SPSS

TL;DR

Run the moderation model twice in SPSS: once without the interaction term and once with the interaction term (CC × role ambiguity).

Briefing Cornell Notes

Briefing

Role ambiguity weakens the positive relationship between collaborative culture and organizational performance, and the moderation effect is statistically significant in SPSS. In the model without the interaction term, collaborative culture and role ambiguity explain 34.8% of the variance in organizational performance (R² = 0.348). When the interaction term (collaborative culture × role ambiguity) is added, explained variance rises to 36.3% (R² = 0.363), an increase of 1.5 percentage points—evidence that the interaction contributes additional explanatory power beyond the main effects.

The interaction term is significant and negative, with a beta of −0.095. The corresponding t statistic is −2.741 and the p value is 0.06, which supports the hypothesis that role ambiguity negatively moderates the CC → organizational performance link. Interpreted plainly: collaborative culture is associated with higher organizational performance when role ambiguity is low, but that benefit diminishes as role ambiguity increases. The moderation slope analysis reinforces this pattern by comparing the steepness of the lines on the interaction plot. The line for low role ambiguity is steeper, meaning increases in collaborative culture translate into larger gains in organizational performance. At high role ambiguity, the line flattens, indicating that raising collaborative culture no longer produces the same improvement in organizational performance.

For reporting, the transcript lays out a practical checklist for moderation results: include R² with and without the interaction term, the path coefficients (including the interaction beta), the test statistics and p value for the interaction, the slope interpretation from the moderation plot, and an effect size for the moderation. The effect size is computed using Cohen’s f² for the incremental contribution of the interaction: f² = (R²_included − R²_excluded) / (1 − R²_included). Using R²_included = 0.363 and R²_excluded = 0.348 yields f² = 0.02.

Using Cohen’s benchmarks (0.02–0.15 small, 0.15–0.35 medium, and above 0.35 large), the moderation effect size is small. The overall takeaway is that role ambiguity matters, but its incremental explanatory impact is modest: it significantly weakens the collaborative culture advantage, yet the additional variance explained by the interaction term is relatively limited. The transcript also notes a common sign-checking pitfall—ensuring the interaction’s negative sign is reflected correctly in the interpretation of which line is steeper on the plot—so the narrative matches the statistical direction.

Cornell Notes

The moderation analysis tests whether role ambiguity changes the strength of the relationship between collaborative culture (CC) and organizational performance (OP). Without the interaction term, the model explains 34.8% of OP variance (R² = 0.348). Adding the interaction increases R² to 36.3% (R² = 0.363), a 1.5 percentage-point gain, indicating the interaction adds explanatory value. The interaction coefficient is negative (β = −0.095) and statistically significant (t = −2.741, p = 0.06), so higher role ambiguity weakens the positive CC → OP relationship. Slope analysis shows a steeper CC effect at low role ambiguity and a flatter effect at high role ambiguity. The moderation effect size is small (f² = 0.02).

Why does the analysis require R² “with” and “without” the interaction term?

R² with the interaction (R²_included) captures variance explained by CC, role ambiguity, and their interaction. R² without the interaction (R²_excluded) captures variance explained by the main effects only. The difference between these two values is then used to compute Cohen’s f², which quantifies how much the moderation term contributes to explaining OP beyond the main effects.

How is the moderation effect interpreted when the interaction beta is negative?

A negative interaction beta (here, β = −0.095) means the moderator weakens the relationship between CC and OP. The main CC→OP association is positive, but as role ambiguity increases, the slope linking CC to OP becomes less steep—so the positive effect of CC on OP diminishes under higher role ambiguity.

What does the slope (interaction plot) reveal, and how should steepness be used?

The moderation plot is interpreted by comparing line steepness at low versus high values of the moderator. In this case, the line is steeper at low role ambiguity: increases in CC lead to larger increases in OP. At high role ambiguity, the line flattens: increases in CC do not translate into comparable gains in OP. This visual pattern matches the negative interaction coefficient.

How is f² for moderation calculated from SPSS output?

Use f² = (R²_included − R²_excluded) / (1 − R²_included). The transcript uses R²_included = 0.363 and R²_excluded = 0.348, giving f² = (0.363 − 0.348) / (1 − 0.363) = 0.02. This f² value is then interpreted using Cohen’s benchmarks for small, medium, and large effects.

What does an f² of 0.02 imply about practical importance?

An f² of 0.02 falls in Cohen’s small effect range (0.02–0.15). So the moderation is statistically meaningful (interaction is significant), but the incremental explanatory impact of role ambiguity on the CC→OP relationship is modest.

Review Questions

What are the specific R² values used to compute the moderation effect size, and what does each represent?
How do the sign of the interaction beta and the slope-plot steepness jointly determine whether the moderator strengthens or weakens the CC→OP relationship?
Using the provided formula, how would you compute f² if R²_included and R²_excluded were different?

Key Points

1
Run the moderation model twice in SPSS: once without the interaction term and once with the interaction term (CC × role ambiguity).
2
Report R² for both models (R²_excluded and R²_included) to show the incremental variance explained by moderation.
3
Use the interaction term’s beta, t statistic, and p value to support or reject the moderation hypothesis.
4
Interpret moderation direction using the sign of the interaction beta: a negative beta means the moderator weakens the main relationship.
5
Use slope analysis by comparing line steepness at low versus high moderator values to describe how CC affects OP under different role ambiguity levels.
6
Compute moderation effect size with Cohen’s f²: (R²_included − R²_excluded) / (1 − R²_included), and interpret it with Cohen’s small/medium/large benchmarks.
7
In this case, role ambiguity significantly weakens the CC → OP relationship, with a small effect size (f² = 0.02).

Highlights

Explained variance rises from R² = 0.348 (no interaction) to R² = 0.363 (with interaction), a 1.5 percentage-point increase from the moderation term.

The interaction is negative and significant (β = −0.095, t = −2.741, p = 0.06), indicating role ambiguity weakens the CC → OP relationship.

Slope analysis shows a steeper CC effect on OP at low role ambiguity and a flatter effect at high role ambiguity.

The moderation effect size is small: f² = 0.02 using Cohen’s incremental contribution formula.

Topics

Moderation Reporting
SPSS Regression
Interaction Effects
Slope Analysis
Cohen's f²