Multilevel Moderation Analysis
Multilevel moderation analysis tests interaction effects in nested data, such as whether a group-level variable changes the strength of an individual-level relationship (a cross-level interaction).
다층 조절분석은 위계적(내재된) 자료에서 집단 수준 변수가 개인 수준 관계의 강도를 바꾸는지(수준 간 상호작용) 등 상호작용 효과를 검정하는 방법이다.
How do I test whether a group-level variable changes an individual-level relationship?
집단 수준 변수가 개인 수준의 관계를 바꾸는지 어떻게 검정하나요?
Multilevel moderation uses a mixed-effects model with a random slope and an interaction between a level-1 predictor and a moderator, often a level-2 variable, to test a cross-level interaction. It fits nested data such as employees in teams or students in schools where relationships may vary across clusters.
When to use Multilevel Moderation Analysis
- Individuals nested in groups, such as students in schools or employees in teams
- Testing whether a team-level variable moderates an individual-level slope
- Repeated measures nested in persons with a person-level moderator
When to use something else
- Data are not nested or clustering is negligible: use ordinary moderation analysis with regression
- The moderated effect runs through a mediator: use multilevel moderated mediation or multilevel SEM
Assumptions and requirements
- Sufficient number of clusters to estimate random slopes
- Level-1 predictors centered within cluster to separate within and between effects
- Normally distributed random effects and residuals
- Correct specification of random effects
What the analysis reports
- Fixed effects including the interaction term with standard errors
- Random intercept and slope variances
- Simple slopes at selected moderator values
- Interaction plot
- Model comparison statistics (AIC, BIC, likelihood ratio test)
How to report it
The cross-level interaction between [level-1 predictor] and [level-2 moderator] was significant, γ = [value], SE = [value], p = [value].
Related methods
- Multilevel Model (Linear Mixed Model)
- Moderation Analysis
- Multilevel Mediation Analysis
Run Multilevel Moderation Analysis on Quorant
- Data you need
- a table (CSV or Excel) with an outcome variable, a cluster ID and predictor (feature) columns
- Price
- 900 credits for one analysis up to 10,000 rows (complexity level 5); larger or wider data costs more. Optional methodologist review: 3600 credits.
- Checking
- Python results are cross-checked against R reference outputs in every deployment
- Before you pay
- A free pre-analysis diagnosis checks your data and gives an exact quote; nothing is deducted until you approve it.