Description Usage Arguments Value Examples
This fits a model estimating the possible hearsay reputation parameters as a multi-group path model given vectors of P1 and P2 reports (vectors of quoted variable names) and a group-level categorical variable. The baseline model estimates each parameter seperately, labelling parameters based on the labels of moderator variable. Those parameters are:
hearsay consensus; the correlation between P1(T) & P2(T)
Intercept for P1(T)
Intercept for P2(T)
variance for P1(T)
variance for P2(T)
P1-P2 Relative Elevation (i.e., Mean P1(T) - Mean P2(T))
If n exchangeable triads > 1:
direct reciprocity; the correlation between opposit P1(T)s (e.g., A(C) <-> C(A))
hearsay reciprocity; the correlation between exchangeable P2(T)s (e.g., B(C) <-> D(A))
unnamed parameter; The correlation between P2(T) and the opposite P1(T) in a group. (e.g., B(C) <-> C(A))
The function can handle up to n exchangeable triads.
1 2 3 4 |
data |
The dataframe that contains P1 & P2 ratings and the group-level moderator. Data should be wide, with a row for every group of participants. At a minimum, it must contain three columns: one for P1 reports, one for P2 reports, and one for the group-level moderator. |
model |
Optional. A model from the corresponding ReputationModelR model builder function. If this is supplied, no additional arguments need to be specified. |
p1_reports |
Quoted column names that contain P1 reports, or ratings made by the person that knows the target directly. If more than one is supplied, the target-wise order must match the other rating types. |
p2_reports |
Quoted column names that contain P2 reports, or ratings made by the person that knows the target indirectly through the corresponding P1. If more than one is supplied, the target-wise order must match the other rating types. |
group_mod |
The quoted column name that contains a group-level categorical moderator. |
use_labs |
Logical indicating whether or not to use the group labels to create the parameter labels. If FALSE, generic labels (grp1 to grpk, where k is the number of groups) are used. |
groups_eql |
Optional. Groups that you want to constrain to be equal across some or all parameters. If you have use_labs set to TRUE, provide a vector of group labels corresponding to the groups you want to constrain to be equal. If you have use_labs set to FALSE, provide a vector of numbers corresponding to the position of the groups you want to constrain to be equal. If you provide "all", all groups will be constrained to be equal. |
params_eql |
Optional. Parameters that you want to constrain to be equal across the groups specified in groups_eql. You can provide one or more specific parameters (e.g., "hc" for hearsay consensus), or use one of several built-in options including "all" which constrains all parameters to be equal across groups and "main" which constrains just the hearsay consensus to be equal across groups (in this model). |
n_triads |
The number of exchangeable triads in each group. By default, this is determined by counting the number of P1 reports. This parameter rarely needs to be changed. |
n_p1s_per_p2s |
The number of P1s for every P2. This defaults to 1. Currently, only values of 1 are supported. |
n_p2s_per_p1s |
The number of P2s for every P1;. This defaults to 1. Currently, only values of 1 are supported. |
The function returns an object of class lavaan
.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 | data("rep_sim_data")
agree_consensus_grpmod <- rep_consensus_group_mod(data = rep_sim_data,
p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
group_mod = "study")
# alternatively
# build the model
agree_consensus_grpmod_model <- rep_consensus_group_mod_builder(p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
groups = levels(rep_sim_data$study))
# then fit it
agree_consensus_grpmod <- rep_consensus_group_mod(data = rep_sim_data,
model = agree_consensus_grpmod_model,
group_mod = "study")
# fit model with group equality constraints:
# if we wanted to constrain all parameters to be equal across the 2 groups, that can be done
# by setting groups_eql and params_eql both to "all".
agree_consensus_grpmod <- rep_consensus_group_mod(data = rep_sim_data,
p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
group_mod = "study",
groups_eql = "all",
params_eql = "all")
# Or we could constrain just hearsay consensus to be equal
agree_consensus_grpmod <- rep_consensus_group_mod(data = rep_sim_data,
p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
group_mod = "study",
groups_eql = "all",
params_eql = c("hc", "v_p1", "v_p2"))
# It can also handle more groups, and groups without unlabelled groups.
# The simulated dataset has the group_var variable, which is contains
# 4 groups, each labelled with just a number (1 to 4). To use an unlabelled group
# either set use_labs to FALSE or it will do so for you.
agree_4grp_consensus_fit <- rep_consensus_group_mod(data = rep_sim_data,
p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
group_mod = "group_var")
# for unlabelled groups (or whenever use_labs = FALSE), you can select certain groups for equality constraints
# by passing a vector of group numbers that should be constrained to be equal. For example, if we wanted
# just groups 1 and 3 to be equal, we would set groups_eql to c(1, 3), like so:
agree_4grp_consensus_fit <- rep_consensus_group_mod(data = rep_sim_data,
p1_reports = c("A_C_agreeableness", "C_A_agreeableness"),
p2_reports = c("B_C_agreeableness", "D_A_agreeableness"),
group_mod = "group_var",
groups_eql = c(1, 3),
params_eql = "all")
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.