Comparing MPlus model solutions via tidyLPA

Models that are specified

All five variables standardized w/ strict defaults (not evaluated)

1. All five variables

out_df1 <- compare_solutions_mplus(df, 
                                   dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                                   starts = c(600, 120),
                                   n_processors = 8)

##   n_profiles                    model_1                    model_2
## 1          2                  39916.157                  38423.266
## 2          3                  39082.592                  38049.877
## 3          4                  38616.439                  37623.057
## 4          5                  37907.604                  37301.328
## 5          6                  37617.262 Warning: LL not replicated
## 6          7                  37182.108                   34517.58
## 7          8 Warning: LL not replicated Warning: LL not replicated
## 8          9 Warning: LL not replicated Warning: LL not replicated
## 9         10 Warning: LL not replicated Warning: LL not replicated
##                    model_3                  model_4
## 1 Error: Convergence issue Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue
##                    model_5                  model_6
## 1 Error: Convergence issue Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue

2. Just three (with composite of engagement)

out_df2 <- compare_solutions_mplus(df, 
                                   dm_overall_eng, dm_challenge, dm_competence,
                                   starts = c(600, 120),
                                   n_processors = 8)

##   n_profiles                    model_1                    model_2
## 1          2                  23932.751                  23020.317
## 2          3                   22766.91                  22636.052
## 3          4                  22177.436 Warning: LL not replicated
## 4          5 Warning: LL not replicated                   21558.95
## 5          6                  21773.331 Warning: LL not replicated
## 6          7                  21191.709                  20692.102
## 7          8                  21792.204 Warning: LL not replicated
## 8          9 Warning: LL not replicated Warning: LL not replicated
## 9         10   Error: Convergence issue Warning: LL not replicated
##                    model_3                  model_4
## 1 Error: Convergence issue Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue
##                    model_5                  model_6
## 1                22850.337 Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue

3. Just engagement

out_df3 <- compare_solutions_mplus(df, 
                                   dm_cog_eng, dm_beh_eng, dm_aff_eng,
                                   starts = c(600, 120),
                                   n_processors = 8)

##   n_profiles                    model_1                    model_2
## 1          2                  23497.708                  22585.734
## 2          3                  22720.718                  22202.609
## 3          4 Warning: LL not replicated                  21707.458
## 4          5                  21709.001                  21433.118
## 5          6                  21412.842                  18561.022
## 6          7                  21379.115 Warning: LL not replicated
## 7          8                  21197.108 Warning: LL not replicated
## 8          9                   20857.06 Warning: LL not replicated
## 9         10 Warning: LL not replicated Warning: LL not replicated
##                    model_3                  model_4
## 1 Error: Convergence issue Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue
##                    model_5                  model_6
## 1 Error: Convergence issue Error: Convergence issue
## 2 Error: Convergence issue Error: Convergence issue
## 3 Error: Convergence issue Error: Convergence issue
## 4 Error: Convergence issue Error: Convergence issue
## 5 Error: Convergence issue Error: Convergence issue
## 6 Error: Convergence issue Error: Convergence issue
## 7 Error: Convergence issue Error: Convergence issue
## 8 Error: Convergence issue Error: Convergence issue
## 9 Error: Convergence issue Error: Convergence issue

1i. Following up on all five variables

m2_4 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 2, n_profiles=4,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 18659.676

## BIC is 37623.057

## Entropy is 0.922

m2_5 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 2, n_profiles=5,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 18474.834

## BIC is 37301.328

## Entropy is 0.901

m2_6 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 2, n_profiles=6,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## Warning: LL not replicated

m2_7 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 2, n_profiles=7,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 17035.006

## BIC is 34517.58

## Entropy is 0.965

m2_8 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 2, n_profiles=6,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## Warning: LL not replicated

## [1] "Warning: LL not replicated"

## [1] "Warning: LL not replicated"

m1_4 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 1, n_profiles=4,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 19196.328

## BIC is 38616.439

## Entropy is 0.811

m1_5 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 1, n_profiles=5,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 18817.934

## BIC is 37907.604

## Entropy is 0.913

m1_6 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 1, n_profiles=6,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 18648.785

## BIC is 37617.262

## Entropy is 0.888

m1_7 <- estimate_profiles_mplus(df, 
                        dm_cog_eng, dm_beh_eng, dm_aff_eng, dm_challenge, dm_competence,
                        starts = c(600, 120),
                        model = 1, n_profiles=7,
                        n_processors = 8)

## Note that this (and other functions that use MPlus) is at the experimental stage! Please provide feedback at https://github.com/jrosen48/tidyLPA

## LogLik is 18407.232

## BIC is 37182.108

## Entropy is 0.886

## # A tibble: 2,958 x 12
##    DM_COG_E DM_BEH_E DM_AFF_E DM_CHALL DM_COMPE  CPROB1  CPROB2  CPROB3
##       <dbl>    <dbl>    <dbl>    <dbl>    <dbl>   <dbl>   <dbl>   <dbl>
##  1     4.00     3.00     3.00     2.00     3.00 0.00100 0       0.0120 
##  2     3.00     3.00     3.00     1.00     4.00 0.00300 0       0.0500 
##  3     3.00     1.00     4.00     1.00     4.00 0       0       0      
##  4     3.00     1.00     3.00     3.00     4.00 0       0.00100 0      
##  5     3.00     3.00     3.00     2.00     3.00 0.00200 0       0.00400
##  6     3.00     4.00     2.00     3.00     4.00 0.896   0       0.00100
##  7     3.00     2.00     4.00     1.00     4.00 0       0       0      
##  8     3.00     3.00     3.00     3.00     4.00 0.00100 0       0.00300
##  9     3.00     3.00     4.00     1.00     4.00 0       0       0.739  
## 10     3.00     3.00     3.00     2.00     3.00 0.00200 0       0.00400
## # ... with 2,948 more rows, and 4 more variables: CPROB4 <dbl>,
## #   CPROB5 <dbl>, CPROB6 <dbl>, C <dbl>

2i. Following up on composite of engagement

## [1] "Warning: LL not replicated"

2i. Following up on just engagement

## [1] "Warning: LL not replicated"

Comparing MPlus model solutions via tidyLPA

Joshua Rosenberg

12/27/2017

Models that are specified

1. All five variables

2. Just three (with composite of engagement)

3. Just engagement

1i. Following up on all five variables

2i. Following up on composite of engagement

2i. Following up on just engagement