Skip to main content
. Author manuscript; available in PMC: 2020 Jul 1.
Published in final edited form as: Addict Behav. 2018 Aug 29;94:74–82. doi: 10.1016/j.addbeh.2018.08.030

Figure 3. Latent variable moderation analysis visualized using small multiples of scatter plots. [Color imagine online only].

Figure 3.

Caption: Scatter plots (top) show latent estimates of X and Y graphed across quartiles of moderator variable Z. The corresponding path diagram is also shown (bottom). Data are simulated for this example. All path coefficients are statistically significant (see supplemental R syntax). See Figure 2 caption for explanation of model fit statistic abbreviations.