Hello,   Using an ecological assessment design (where people report their negative affect and social interaction every three hours each day across 5 days of the week), I am trying to calculate's participants' hourly experience of Negative Affect (NA) when interacting with family members compared to No interaction (FamilyInteraction: categorical var).   I am trying to see how trait levels of neuroticism and extraversion may further explain this relationship so I am using a  three-way interaction with a dichotomous variable (FamilyInteraction) and two continuous variables (neuroticism and extraversion).   So far, the only way I have figured out a way to calculate the simple slopes and slope contrast is by using proc plm and using the lsmestimate command. But in order to use this syntax, I have to make one of the personality vars into a categorical variable. If I want to keep both of my personality variables as continuous variables, what syntax code can I use?   *Below purpose syntax is the main model of the analysis proc mixed noclprint covtest ; class ID FamilyInteraction; model NA= FamilyInteraction|Neuro|Extraversion/ solution ddfm = bw; RANDOM intercept / SUBJECT=Day; RANDOM intercept / SUBJECT=ID(Hour); store Save;run; proc plm noclprint restore=Alonea; effectplot interaction (x=FamilyInteractionsliceby=Neuro) / at(Extra=1.0741531 1.9140935) clm connect; run;   *Below syntax in blue the way I would calculate the slopes & slope differences if one of the personality traits (Neuro) is made into a categorical variable and used in the proc plm lsmestimate command. proc plm noclprint restore=Alonea; lsmestimate C1r*LonelyCat 'No Fam-Yes Fam, Low Neuro, Extra=Low' [1, 1 1] [-1, 2 1],                                              'No Fam-Yes Fam, High Neuro, Extra=Low' [1, 1 2] [-1, 2 2] / e adj=bon at  Extra=1.0741531; lsmestimate C1r*LonelyCat 'No Fam-Yes Fam, Low Neuro, Extra=High' [1, 1 1] [-1, 2 1],                                              'No Fam-Yes Fam, High Neuro, Extra=High' [1, 1 2] [-1, 2 2] / e adj=bon at Extra=1.9140935; run; proc plm noclprint restore=Alonea; lsmestimate FamilyInteraction*Neuro' diff No Fam-Yes Fam, LowNeuro- HighNeuro, Extra=Low' [1, 1 1] [-1, 2 1] [-1, 1 2] [1, 2 2] / e at Extra=1.0741531 joint; lsmestimate FamilyInteraction*Neuro' diff No Fam-Yes Fam, LowNeuro- HighNeuro, Extra=High' [1, 1 1] [-1, 2 1] [-1, 1 2] [1, 2 2] / e at Extra=1.9140935 joint; run;   I really appreciate any help I can get and please let me know if I can further clarify anything! ... View more

Hello,   I am trying to merge one wide and one long data set. Both datasets have information on the same participants which is identified by the variable ID.    Once I sort the datasets by ID, can I merge them from their original data formats? Or do I have to make both datasets into the same format (both long or both wide files) in order to merge them?   I greatly appreciate your help!   Best, Joanna ... View more

I’m looking at intra-individual variability for three variables (rate life in the past, present and the future) using the Proc Mixed command in SAS. My main predictors are Sample( nationality:Japan vs. U.S.), 5 Age groups and three time points (past, present and future). One of my level 2 variables(extraversion personality trait, which I am adjusting as a covariate) correlates very differently with the three main variables. My model only works when I also include this level 2 variable (Extraversion) in the random statement. I know we only put level 1 variables (slopes) in the random statements and not level 2 variables so I am not sure why the model only works when I put the level 2 variable in the random statement.   proc mixed noclprint covtest ; class sample AgeGroup5 time marriedyn gender ; model ratelife = gender NewChronNumb marriedyn extraversionnew|time NeuroticismNew ZEduHighest sample|time|AgeGroup5 /solution ddfm=bw; random intercept extraversionnew / type=un subject=id; lsmeans sample*time*agegroup5; run;     Thank you for your help in advance! Joanna ... View more