How should two categorical independent variables with three levels and their interactions be handled in a linear regression in SPSS?

17 May 2023 2 991 Report

I am doing my thesis and in my design I have two independent variables, both with 3 levels (no frame, low and high). In the model I am also using the two-way interaction between these two and the two-way and three-way interaction with moderators (party preference and political trust). I found some tutorials online that say I should code the independent variables as a set of two variables to use both in the regression model (D1: no frame (0), low (1), high (0) and D2: no frame (0), low (0) and high (1)). However, when I enter D1.1 and D1.2 for independent variable 1 and D2.1 and D2.2 for independent variable 2 into the model, SPSS exlcudes one of the variables in the output (D1.1).

Moreover, I am not sure how to code and add the interactions to the model. Initially, I coded one variable for each independent variable as follows: no frame (-1), low (0), high (1) and used these for the interactions. Is that also an option, or does this cause problems?

David Morse

Hello Bieke,

If you genuinely believe that the unit difference separating "no frame" and "low" is exactly the same as that separating "low" and "high" for an IV, then yes, you may code them as -1, 0, 1 (or any other linear transformation thereof). However, if you don't believe this, then you must opt for recoding, replacing a three-level variable with two dummy variates.

I suspect that the reason one of your variables is being dropped is due to redundancy, and this could be caused by: (a) no instances of one of the three categories in the data set; (b) IV1 and IV2 are redundant in some way (e.g., every case is in the same level for both variables); (c) a mistaken recoding.

Are you also including all four of the dummy variate products to capture the information about their interaction?

Good luck with your work.

Ma'Mon Abu Hammad

When dealing with two categorical independent variables with three levels each and their interactions in a linear regression analysis in SPSS, you can follow these steps:

Open your dataset in SPSS or create a new dataset.

In the "Variable View" tab, define your independent variables as categorical variables and assign appropriate value labels to each level. For example, if you have two categorical variables named "FactorA" and "FactorB," each with three levels (e.g., Low, Medium, High), you would specify them as categorical variables and assign value labels accordingly.

Go to "Analyze" in the menu bar, then select "General Linear Model" and choose "Univariate."

In the "GLM Univariate" dialog box, select your dependent variable and move it to the "Dependent Variable" box.

Add the two categorical independent variables (FactorA and FactorB) to the "Fixed Factors" box.

To include the interaction effect between FactorA and FactorB, click on the "Model" button. In the "Model" dialog box, select the "Full factorial" option under "Custom Model" and move the interaction term (FactorA * FactorB) to the "Terms" box.

Customize any additional settings, such as post hoc tests or planned contrasts, if applicable.

Click "OK" to run the analysis.

SPSS will generate the output, which includes information about the main effects of FactorA and FactorB, the interaction effect between FactorA and FactorB, as well as other relevant statistics such as coefficients, p-values, and effect sizes.

It's important to interpret the results carefully, considering the significance of the main effects and interaction effect, and considering any post hoc tests or planned contrasts you may have conducted.

Remember to evaluate the assumptions of linear regression, such as linearity, independence, and homoscedasticity, and assess model fit using appropriate statistics, such as R-squared or adjusted R-squared.

Additionally, if you encounter issues related to multicollinearity or small cell sizes within your categorical variables, you may need to consider alternative modeling techniques or data transformations to address these concerns.

How to compare two groups with only two measurements?

What's the best way to measure growth rates in House sparrow chicks from day 2 to day 10?

After a reference from Int J Body Compos Res, can any body help?

Logistical regression using SPSS for a dichotomous dependent variable, with an independent variable of 1, 2, 5 and 10 seconds?

What techniques can I use to determine the miscibility of polymer blends?

Do I need to do linear regression before running mediation analysis using PROCESS MACRO by Hayes ?

How to code open-ended questions in SPSS?

How to repeat an operation on an atom selection for multiple times in PyMol?

Shall I skip the measurement model in SmartPLS?

What is involved in morphological changes of SEM images of a substance?