growcurves (version 0.2.3.6)

datsimmult: Repeated measures for two groups of subjects with two multiple membership (MM) terms

Description

A simulation dataset containing repeated subject measures for 2 treatment arms, (control = 0, treatment = 1), constructed from a four- MM term multiple membership model. MM terms 1 and 3 employ an 'mmi' prior over the set of treatment groups, which are then linked back to subjects with a weight matrix. MM terms 2 and 4 employ an 'mmcar' prior over the number of treatment sessions attended, with correlations between adjacent number of sessions equal to 0.25. Subject effects were randomly drawn from 10 clusters with assignment weights/probabilities drawn from a Dirichlet distribution. Cluster location values were generated from a Gaussian base distribution.

format

A list object of 22 variables for 897 total observations on 299 subjects

Details

• y. response. there areN = 897total measures forP = 299subjects
• subject. subject identifier(1,2,...,299
• trt. treatment group identifier of lengthN(e.g.(0,0,0,...,1,1,1,...), either{0,1}for control and treatment.
• time. times in months for each repeated subject measure of lengthN. There are 3 distinct time points. e.g.(0,3,6,0,3,6,0,0,3,,,,)
• n.random. number of random effects per subject. Set = 3.
• n.fix_degree. order of fixed effects. Set = 2, for quadratic, meaning 3 effects (intercept, slope, quadratic) each, for treatment and control groups.
• X The resultantN x n.fixedfixed effects design matrix.X = c(1, time, time^2, trt_1,trt_1*time, trt_1*time^2, age, income).
• coefs. true coefficient values for the time-based quadratic fixed effects generated from thetrt, andtimeinputs.
• Z The resultantN x n.randomby-subject random effects design matrix.Z = c(1, time, time^2.
• Z.b TheN x 1matrix [Z_{1} * b_{1},...,Z_{n}*b_{n}]. wherenequals the number of (unique) subjects.
• subj.aff. A list object with each term a vector that indexesn[g]subjects linked to each ofg = 1, ..., (G = 4)CBT therapy groups. Each group is specialized to its own multiple membership (MM) term for employment of engine functiondpgrowmmult. The number of treatment arm subjects for each group are(n[1] = 17, n[2] = 18, n[3] = 19, n[4] = 78)for a total ofPaff = 132subjects assigned to the treatment arm.
• W.subj.aff. A list object containingG, n[g] x S[g]multiple membership weight matrices that together map theP_aff = 132affected subjects (in each element of thesubj.afflist) to their particular sessions attended within their assigned group. There are a total of 245 sessions allocated to theGgroups asS[1] = 36, S[2] = 40, S[3] = 40, S[4] = 129.
• group. A list object of length equal to the number of MM terms under the 'mmcar' prior - In this case, terms 2 and 4. The elements in this list object contain a vector of labels that align the MM effects in a given term to disjoint, non-communicating groups. For these data, there are no groupings, so the single entry in 'group' associated to the 'mmcar' prior on number of sessions attended is filled with 1's equal to the number of of MM effects in each term.
• Omega. A list object containing anS[g] x S[g]CAR adjacency matrix used to model prior dependence among effects for each MM term under the 'mmcar' prior, whereS[g]are the number of effects for termi. Here, there are two terms (2 and 4) under "mmcar", soOmegacontains two elements.
• us. A list object where each element is a vector of true MM session effect values for a given MM term. There are four MM terms in these data, so there are 4 elements, each of lengthS[g].
• s. A vector of true cluster memberships for each of thensubjects.
• b.star. A list object of true cluster location values for each ofM = 10clusters. Each entry contains then.random = 3location values for that cluster.
• b. A list object true random effect coefficient values for each ofnsubjects. Each entry contains then.random = 3effect values for that subject.
• coefs. A numeric vector of length {n.fixed + 1} that contains the true fixed effects coefficient values for the intercept plus the columns inX(in order).
• tau.b. true values for the prior precisions of the base Gaussian distribution for each ofn.random = 3subject effects.
• tau.u. An4 x 1vector of precision parameters associated to the Gaussian prior formulations for each MM term (under either 'mmcar' or 'mmi' covariance constructions).
• tau.e. true value for overall model error.
• option. A character vector equal toc("mmi","mmcar","mmi","mmcar")that presents the prior formulations chosen for the 4 MM terms; The "mmi" employs an independent Gaussian prior while the "mmcar" allows for adjacency based dependence.