cleaning up equations

Serial Mediation.md

@@ -12,11 +12,16 @@ A--> |c'| C;
 ```
 
 $B1 = \beta_0 + a_1 \cdot A + e_i$
+
 $B2 = \beta_0 + a_2 \cdot A + b_1 \cdot B1 + e_i$
+
 $C=\beta_0 + c' \cdot A + b_3 \cdot B1 + b_2 \cdot B2 + e_i$
+
 ## Pathways
 $a_1 \cdot b_1 \cdot b_2$
+
 $a_2 \cdot b_2$
+
 $a_1 \cdot b_3$
 
 ---

Simple Moderation.md

@@ -12,8 +12,11 @@ W --> D{ };
 W --> |m| C;
 ```
 $C = \beta_0 + c\cdot A + m\cdot W + b\cdot A\cdot W$
+
 This is the regression equation for this model. Which can be rewriteen as:
+
 $C = \beta_0 + m\cdot W + (c + b\cdot W)\cdot A$
+
 When written like this you can see how the A variable is now a linear function of the moderating variable. 
 
 To interpret this model first determine if parameter $b$ is significant. Then the model can be [[Probing an Interaction|probed]] to identify the range of values of W for which the A effect is significant. 

The direct effects matrix.md

@@ -21,10 +21,11 @@ A-->|c'| C;
 |C|1|1|0|
 
 This matrix is read row by row with the variable in each row being the outcome variable. Each column with a value in it is a predictor in the model. In the above figure, there are two equations. The two equations are:
+
 $B = \beta_0 + \beta_1 \cdot A + e_i$
+
 $C = \beta_0 + \beta_1 \cdot A + \beta_2 \cdot B + e_i$
-
-
+
 
 It is also possible that an analysis requires the same outcome variable in multiple equations. Even in the simplest model this is the case. Therefore, there needs to be a second layer (third dimension) to the above matrix. It will be like this: