Article directory
1. First of all, you must understand what kind of adjustment and independent variables you have
2. Practical exercises
In this exercise, we take the example that the independent variable is a dichotomous variable and the moderating variable is a continuous variable and a categorical variable respectively, to teach you to test the moderating effect.
Note:
1. When the independent variable is a dichotomous variable, no centering is required.
2. The adjustment variable must be centered first, and then the product term with the independent variable is calculated.
2.1 When the adjusted variable is a continuous variable
For example, the level of self-monitoring below is a continuous variable, and the style of cuteness is a dichotomous variable to test for moderating effects.
In my data, the independent variable is xiushi, the adjustment variable is ml, and the mediator is adh.
2.1.1 Centralize ml
Talk about the independent variable *ml centralization=interaction item directly in the calculation variable of spss (this step is very simple, if you don’t understand, you can leave a message, if many people don’t understand, I am making a tutorial)
2.1.2 Use hierarchical regression to explore the influence of the interaction of independent variables and ml on adh
At the first level, we put the control variables in,Open spss-analysis-regression-linear.
In the second layer, we put the independent variables in.
In the third layer, we put the central adjustment variables in.
In the fourth layer, we put the interaction items in.
Finally, check these few in the statistics side, and you can analyzed
2.1.3 Interpretation of results
- In the result table in model 4, its significance is established, indicating that our adjustment and independent variables have interaction effects. Then we can first report the F value below. For example, in my research, F(10, 273)=7.740, p is less than 0.001,
- Find model 4 in this coefficient table (mosaic is used here for data confidentiality)
The last three lines are the independent variable, the central adjustment variable, and the interaction item. The three values in the red box are very important.
The significance of the interaction term of the circle is also lower than 0.05, which also shows that the interaction is significant.
- We need to find this table dedicated to simple slope analysis, and paste these values into it.
4) Further down, we only need to fill in the standard deviation of the independent variable and the adjusted variable. This can be calculated in the analysis description in spss, and pasted in.
5) Next we need to fill in this part, we need to find the coefficient correlation table in the regression analysis just now,
In Excel, the variance of coefficient of IV is the covariance of the independent variable to the independent variable in the coefficient table. The
Variance of coefficient of interacting is the interaction item to the interaction item.
The next line is the covariance of the independent variable to the interaction item
. Therefore, everyone in this coefficient correlation Just find these values and paste them in.
Sample size is the sample size and
Number of control variable is the control variable, just fill in these two.
- Then you can get a simple slope plot and the following high and low effect values
B here should be the value of gradient
Next update:
"2.2 Test method when the moderating variable is a categorical variable"
If you like it, you can bookmark and like it,,,
your likes are the motivation for updating