|
|
Testing causality with multi-level data |
|
Message/Author |
|
|
Hello, I want to test the causal direction of a relationship between a group-level variable (nation-level scores) and an individual-level variable. In particular, in a multi-country sample I want to test whether culturally-shared values (level-2 variable) impact attitudes of individuals, or whether, instead, individual attitudes impact culture-level values. Is there a technique to do that with MPlus? Best regards David |
|
|
The only way to have an group-level variable influence and individual-level variable is it it influences the group part of the individual-level variables, for example, %WITHIN% y; %BETWEEN% y ON w; where y is an individual-level variable and w is a group-level variable. |
|
Back to top |
|
|