Basic question about Multilevel Analysis and MlWiN: Group vs. grand mean centering
Posted: Thu Dec 19, 2019 12:16 pm
Hello everyone,,
I am currently doing my bachelor thesis in Psychology and have to make a multilevel analysis. Basic Info: Im looking into the relationship between time pressure at work and the risk of sport injuries in trainings after work. Over the course of a few months we gathered data from about 60 individuals in between 3 and 5 trainings. There was a questionnaire about work before the training, which assessed time pressure and other work demands and cognitive failures. A second questionnaire was given to the individuals asking them about near accidents and accidents. Another questionnaire was given to them at the beginning of the study to measure personality variables, the overall time pressure, perceived work control and such. As discussed with my instructor we'll use MM-Analysis to test our hypotheses. Now my question: We defined two levels: Level 1 being the measures at the day to day training level, such as time pressure at work, amount of near injuries and injuries, etc. Level 2 consists of the person variables.
I know that the level 2 predictors have to be grand mean centered, and my instructor advised me to group mean center the Level 1 predictors. Im not entirely sure as to what the difference is and wanted to ask if someone could explain this to me and why its better to use the group mean centering at level 1.
Thanks for your help and greetings from Switzerland.
I am currently doing my bachelor thesis in Psychology and have to make a multilevel analysis. Basic Info: Im looking into the relationship between time pressure at work and the risk of sport injuries in trainings after work. Over the course of a few months we gathered data from about 60 individuals in between 3 and 5 trainings. There was a questionnaire about work before the training, which assessed time pressure and other work demands and cognitive failures. A second questionnaire was given to the individuals asking them about near accidents and accidents. Another questionnaire was given to them at the beginning of the study to measure personality variables, the overall time pressure, perceived work control and such. As discussed with my instructor we'll use MM-Analysis to test our hypotheses. Now my question: We defined two levels: Level 1 being the measures at the day to day training level, such as time pressure at work, amount of near injuries and injuries, etc. Level 2 consists of the person variables.
I know that the level 2 predictors have to be grand mean centered, and my instructor advised me to group mean center the Level 1 predictors. Im not entirely sure as to what the difference is and wanted to ask if someone could explain this to me and why its better to use the group mean centering at level 1.
Thanks for your help and greetings from Switzerland.