# SPSS Statistics

View Only

## Need help to run a mixed linear regression

• #### 1.  Need help to run a mixed linear regression

Posted Wed May 05, 2021 10:03 AM
Hello everyone,

I am is the process of analysing data for my master thesis and I am encountering problems to run a mixed linear regression.
I am a beginner with SPSS statistics and despite watching many videos and reading tutorial I do not manage to apply it to my data.

My conceptual model is the following:
- IV: Level of Self-Esteem (continuous variable that I mean centered)
- DV: Likelihood to complain after a negative service experience (measured on a 0-5 scale)
- Moderator: Expression mode to complain (face-to-face or email)

Likelihood to complain has been measured in two different situations: via Face-to-face or via email.
I put my data in long format so each respondents has two rows with likelihood to complain in situation 1 or 2 (i actually have effect coded dummies -1 and 1 for situation 1 (face-to-face) or 2 (email) - and level of self-esteem of the respondents.

I am trying to run an analysis to see main effect of the IV, main effect of the moderator and interaction effect (so I created an interaction term).

I do not understand which of my variables is random effect and/or fixed effect?
Should I also be mentioning any variable in the first window of the analysis (subjects and repeated)?

The reason why I want to conduct a mixed linear regression is because my supervisor said so, but it is all completely new to me.

Thank you in advance to anyone who could help me even a little bit! If I need to provide any data file please let me know.

------------------------------
Ophélie M.
------------------------------

• #### 2.  RE: Need help to run a mixed linear regression

Posted 2 days ago
https://towardsdatascience.com/how-linear-mixed-model-works-350950a82911

------------------------------
Mrunang Rathod
------------------------------

• #### 3.  RE: Need help to run a mixed linear regression

Posted 2 days ago
Here is a link (non-IBM) to a webinar that addresses fixed vs random factors in mixed models.  Maybe it will help you decide.

There are downloadable slides as well as the recording of the webinar.

--