To construct a Structural Equation Model (SEM) in AMOS with chain mediators, follow these
steps:
### Example Model Layout with Chain Mediators
1. **Define Your Variables:**
- **Independent Variable (IV):** Ostracism
- **Dependent Variable (DV):** Gaming Addiction
- **Mediators:** Fear of Missing Out (FOMO) and Self-Esteem
### Steps to Construct the SEM
1. **Open AMOS:**
- Start AMOS Graphics.
2. **Draw Your Model:**
- **Create Variables:**
- Use the "Draw observed variables" tool to draw rectangles for each observed variable
(Ostracism, FOMO, Self-Esteem, and Gaming Addiction).
- **Add Arrows:**
- Use the "Draw paths" tool to draw arrows from:
- Ostracism to FOMO
- FOMO to Self-Esteem
- Self-Esteem to Gaming Addiction
- **Add Error Terms:**
- Use the "Draw unobserved variables" tool to draw circles for error terms for each dependent
variable and mediators. Connect these error terms to their respective variables.
3. **Specify Relationships:**
- Double-click on each path to specify the regression weights.
- Ensure that each path reflects the direction of the hypothesized relationships.
4. **Add Covariances (if needed):**
- If there are any covariances between the error terms of the mediators, draw double-headed
arrows to represent these covariances.
5. **Label Your Variables:**
- Label each variable and error term clearly to avoid confusion.
6. **Define the Model in AMOS:**
- Click on "Analyze" and then "Calculate Estimates" to define and run your model.
### Diagram Illustration
Here's a simplified visual layout with chain mediators:
```
Ostracism --> FOMO --> Self-Esteem --> Gaming Addiction
```
### Running the Analysis
1. **Data Input:**
- Ensure your data is correctly entered in an SPSS file and linked to AMOS.
2. **Model Specification:**
- Click on "Analyze" and specify the type of analysis (e.g., SEM).
3. **Run the Model:**
- Click "Calculate Estimates" to run the model and obtain the results.
4. **Check Model Fit:**
- Review fit indices such as CFI, TLI, RMSEA, and Chi-square to evaluate model fit.
5. **Interpret Results:**
- Interpret the path coefficients, direct and indirect effects, and overall model fit.
By following these steps, you should be able to construct and analyze your SEM in AMOS with
chain mediators effectively.
In the context of Structural Equation Modeling (SEM), covariances refer to the measure of the
degree to which two variables change together. When you include covariances in your model,
you're indicating that there is a relationship or an association between the variables or error
terms, even if there is no direct causal path specified between them.
### Covariances in SEM
1. **Observed Variables:**
- Covariances between observed variables indicate that you expect a relationship between these
variables. For example, if you expect that FOMO and Self-Esteem are related but do not want to
specify a direct causal path between them, you can include a covariance.
2. **Error Terms:**
- Covariances between error terms of dependent variables or mediators indicate that there is a
relationship between the unmeasured factors influencing these variables. This is common in
models where you suspect that the same unmeasured factors may be affecting multiple variables.
### When to Include Covariances
- **Theoretical Justification:** Include covariances if there is a theoretical reason to believe that
the variables are related.
- **Model Fit Improvement:** Sometimes, adding covariances between error terms can improve
model fit.
- **Correlated Disturbances:** If disturbances (errors) are thought to be correlated due to
unmeasured influences.
### How to Add Covariances in AMOS
1. **Draw Covariances:**
- Use the "Draw covariances" tool (represented by a double-headed arrow) to draw lines
between variables or error terms that you believe are correlated.
2. **Specify Covariances:**
- Double-click on the covariance path to specify the covariance value if you have one, or let
AMOS estimate it.
### Example Diagram
In the context of your model, you might have:
```
Ostracism --> FOMO --> Self-Esteem --> Gaming Addiction
```
If you hypothesize that FOMO and Self-Esteem are correlated, but you do not specify a direct
path, you would add a covariance:
```
Ostracism --> FOMO <--> Self-Esteem --> Gaming Addiction
```
Here, `<-->` represents a covariance.
### Practical Example
1. **Open AMOS:**
- Start AMOS Graphics.
2. **Draw Covariances:**
- Select the "Draw covariances" tool (double-headed arrow).
- Click on the first variable or error term, then click on the second variable or error term you
want to correlate.
3. **Run the Analysis:**
- Click "Calculate Estimates" to run the model and check the covariances among the variables.
4. **Interpret Covariances:**
- Look at the covariance estimates in the output to understand the relationship between the
variables.
Including covariances can help refine your model to better reflect the underlying relationships
among variables and improve the overall fit.
Specifying relationships in AMOS involves defining the paths and associations between
variables in your Structural Equation Model (SEM). This includes setting up regression paths,
covariances, and indicating which variables are independent and dependent.
### Steps to Specify Relationships in AMOS
1. **Drawing the Model:**
- **Independent Variables (IVs):** Draw a rectangle for each IV.
- **Dependent Variables (DVs):** Draw a rectangle for each DV.
- **Mediators:** Draw rectangles for each mediator variable.
2. **Adding Paths:**
- **Regression Paths:** Use the "Draw paths" tool (an arrow) to indicate the direction of
influence from one variable to another. For example, if Ostracism influences FOMO, draw an
arrow from Ostracism to FOMO.
3. **Setting Path Coefficients:**
- After drawing a path, you need to specify the relationship by defining the path coefficients.
This can be done by double-clicking the path arrow, which opens the "Parameters" dialog where
you can input values or constraints.
### Detailed Example
Let's construct the model step-by-step:
#### Example: Ostracism -> FOMO -> Self-Esteem -> Gaming Addiction
1. **Draw the Model:**
- Open AMOS Graphics.
- Draw rectangles for Ostracism, FOMO, Self-Esteem, and Gaming Addiction.
2. **Add Paths:**
- Draw an arrow from Ostracism to FOMO.
- Draw an arrow from FOMO to Self-Esteem.
- Draw an arrow from Self-Esteem to Gaming Addiction.
3. **Add Error Terms:**
- Draw circles for error terms for FOMO, Self-Esteem, and Gaming Addiction. Connect these
circles to their respective variables.
4. **Add Covariances (if needed):**
- If you expect covariances between error terms, draw double-headed arrows between the error
terms.
5. **Label Variables and Paths:**
- Clearly label each variable and path for easy identification.
6. **Specify Path Coefficients:**
- Double-click on each arrow to open the "Parameters" dialog.
- Enter values for the path coefficients if you have them, or let AMOS estimate them.
### Example Steps in AMOS
1. **Draw Variables:**
- Click the "Draw observed variables" icon.
- Click on the canvas to create rectangles for each variable: Ostracism, FOMO, Self-Esteem,
and Gaming Addiction.
2. **Draw Paths:**
- Click the "Draw paths" icon (an arrow).
- Click on Ostracism and drag an arrow to FOMO.
- Click on FOMO and drag an arrow to Self-Esteem.
- Click on Self-Esteem and drag an arrow to Gaming Addiction.
3. **Add Error Terms:**
- Click the "Draw unobserved variables" icon (a circle).
- Click next to each of FOMO, Self-Esteem, and Gaming Addiction to add error terms.
- Connect each error term to its corresponding variable.
4. **Add Covariances:**
- Click the "Draw covariances" icon (a double-headed arrow).
- Click on the first error term and drag to the second error term to add a covariance (if
necessary).
5. **Specify Path Coefficients:**
- Double-click on each arrow to open the "Parameters" dialog.
- Enter the initial values or constraints for the path coefficients, or let AMOS estimate them.
6. **Run the Model:**
- Click "Analyze" and then "Calculate Estimates" to run the model and obtain results.
7. **Check Model Fit:**
- Review fit indices like CFI, TLI, RMSEA, and Chi-square to evaluate the model fit.
### Visual Diagram
Here’s a visual representation:
```
Ostracism --> FOMO --> Self-Esteem --> Gaming Addiction
```
Each arrow represents a regression path, indicating a specified relationship.
### Interpretation
- **Path Coefficients:** Indicate the strength and direction of the relationships between
variables.
- **Model Fit Indices:** Evaluate how well the model fits the data.
By specifying relationships in this manner, you can effectively construct and analyze your SEM
in AMOS, ensuring that your model accurately reflects the hypothesized relationships among
variables.
