Dyadic daily analysis indistinguishable PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Marie-Pier Vaillancourt-Morel posted on Monday, November 26, 2018 - 4:26 am
I am looking for a way of analysis dyadic daily data of indistinguishable parters in Mplus. So both partners of same-sex couples answered to questionnaires everyday for 35 days: days are nested within partners that are nested within couples, but I want to analyse them as a two-level structure (days nested in couples with partners repeated) as both partners answered to the questionnaire on the same day so are kind of crossed (incorporates dependency in error between dyad members at eat time point and within dyad members across time). My analysis will be simple for now, the association between Xactor and XPartner on Yactor (actor and partner associations). I know how to do that with distinguishable dyads in Mplus using both partners on the same line (wide), but I am having a hard time finding how to do that with indistinguishable partners in Mplus as I can not really attribuate number to each partner randomly. I did not find this example in the UG, but it may be too specific.

Any thought or informations/readings would be very appreciate!
 Bengt O. Muthen posted on Monday, November 26, 2018 - 12:57 pm
You can do it as 2-level or 3-level. In both cases, level 1 would be time. With 2-level you would have the partners in wide form on level 2 with equalities of parameters across partners to make them statistically equivalent. With 3-level those equalities would happen automatically. With equalities, I don't think the order of partners in the data makes a difference. I think the same issue occurs in twin modeling although there I think the convention is to have number 1 be for the first-born.
 Marie-Pier Vaillancourt-Morel posted on Monday, November 26, 2018 - 1:05 pm
Thanks a lot for this quick answer. I will try both 2-level and 3-level just to see if it makes some important differences.
 Bengt O. Muthen posted on Monday, November 26, 2018 - 2:22 pm
You should get the same results - it's a good check that you did the equalities right (make sure the number of parameters is the same).
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: