Procedure Steps | Aims | Mplus code included in the.inp syntax file | Translation |
---|---|---|---|
Setup | - Defining the analysis title - Loading the Data -Naming the variables -Defining the nature of the variables, if CATEGORICAL - Defining missing values - The variables to be used in the analyses are also required to be defined - The analysis’ features are then required to be defined | # chosen title follows the command Title: Title: SDQ ESEM 5 factor model for time 1 data # data to be analysed follows the command DATA: File is DATA: file is data.csv; # variable names are provided after the command VARIABLE: Names ARE VARIABLE: Names ARE s1_1 s2_1 s3_1 s4_1 s5_1 s6_1 s7_1 s8_1 s9_1 s10_1 s11_1 s12_1 s13_1 s14_1 s15_1 s16_1 s17_1 s18_1 s19_1 s20_1 s21_1 s22_1 s23_1 s24_1 s25_1; # Categorical variables are provided after the command CATEGORICAL ARE: CATEGORICAL ARE s1_1 s2_1 s3_1 s4_1 s5_1 s6_1 s7_1 s8_1 s9_1 s10_1 s11_1 s12_1 s13_1 s14_1 s15_1 s16_1 s17_1 s18_1 s19_1 s20_1 s21_1 s22_1 s23_1 s24_1 s25_1; # The character(s) defining missing values are provided after the command MISSING ARE all: MISSING ARE all (-9); # The variables to be used in the analyses are provided after the command Usevariable are: Usevariable are s1_1 s2_1 s3_1 s4_1 s5_1 s6_1 s7_1 s8_1 s9_1 s10_1 s11_1 s12_1 s13_1 s14_1 s15_1 s16_1 s17_1 s18_1 s19_1 s20_1 s21_1 s22_1 s23_1 s24_1 s25_1; # The analysis’ features are then selected. After the command ANALYSIS, the type of estimator and rotation are provided via the commands ESTIMATOR IS and ROTATION = respectively ANALYSIS: ESTIMATOR IS wlsmv; ROTATION = TARGET; | The initial Mplus setup involves: a) defining the title of analyses; b) loading the data to be used; c) naming the variables included in the data; d) identifying “categorical” variables within the data; e) providing the missing values’ identifier; g) identifying the specific data variavbles to be used in the analyses and; h) definining the analyses’ estimator and rotation type |
Step 1 | Model setup | # The analysis’ CFA model is defined after the command MODEL: The latent factors are on the left side followed by “BY” indicating the items allocated to them. All non-prmary items are followed by ~ 0, which requests their loadings to be modelled when exceeding a level approximating 0 (this is the exploratory part of the analyses). The last item for each latent factor is fixed (*1) MODEL: PP BY s6_1 s11_1 s14_1 s19_1 s23_1 s1_1 ~ 0 s2_1 ~ 0 s3_1 ~ 0 s4_1 ~ 0 s5_1 ~ 0 s7_1 ~ 0 s8_1 ~ 0 s9_1 ~ 0 s10_1 ~ 0 s12_1 ~ 0 s13_1 ~ 0 s15_1 ~ 0 s16_1 ~ 0 s17_1 ~ 0 s18_1 ~ 0 s20_1 ~ 0 s21_1 ~ 0 s22_1 ~ 0 s24_1 ~ 0 s25_1 ~ 0(*1); CP BY s5_1 s7_1 s12_1 s18_1 s22_1 s1_1 ~ 0 s2_1 ~ 0 s3_1 ~ 0 s4_1 ~ 0 s6_1 ~ 0 s8_1 ~ 0 s9_1 ~ 0 s10_1 ~ 0 s11_1 ~ 0 s13_1 ~ 0 s14_1 ~ 0 s15_1 ~ 0 s16_1 ~ 0 s17_1 ~ 0 s19_1 ~ 0 s20_1 ~ 0 s21_1 ~ 0 s23_1 ~ 0 s24_1 ~ 0 s25_1 ~ 0(*1); ES BY s3_1 s8_1 s13_1 s16_1 s24_1 s1_1 ~ 0 s2_1 ~ 0 s4_1 ~ 0 s5_1 ~ 0 s6_1 ~ 0 s7_1 ~ 0 s9_1 ~ 0 s10_1 ~ 0 s11_1 ~ 0 s12_1 ~ 0 s14_1 ~ 0 s15_1 ~ 0 s17_1 ~ 0 s18_1 ~ 0 s19_1 ~ 0 s20_1 ~ 0 s21_1 ~ 0 s22_1 ~ 0 s23_1 ~ 0 s25_1 ~ 0(*1); HA BY s2_1 s10_1 s15_1 s21_1 s25_1 s1_1 ~ 0 s3_1 ~ 0 s4_1 ~ 0 s5_1 ~ 0 s6_1 ~ 0 s7_1 ~ 0 s8_1 ~ 0 s9_1 ~ 0 s11_1 ~ 0 s12_1 ~ 0 s13_1 ~ 0 s14_1 ~ 0 s16_1 ~ 0 s17_1 ~ 0 s18_1 ~ 0 s19_1 ~ 0 s20_1 ~ 0 s22_1 ~ 0 s23_1 ~ 0 s24_1 ~ 0(*1); PS BY s1_1 s4_1 s9_1 s17_1 s20_1 s2_1 ~ 0 s3_1 ~ 0 s5_1 ~ 0 s6_1 ~ 0 s7_1 ~ 0 s8_1 ~ 0 s10_1 ~ 0 s11_1 ~ 0 s12_1 ~ 0 s13_1 ~ 0 s14_1 ~ 0 s15_1 ~ 0 s16_1 ~ 0 s18_1 ~ 0 s19_1 ~ 0 s21_1 ~ 0 s22_1 ~ 0 s23_1 ~ 0 s24_1 ~ 0 s25_1 ~ 0(*1); | This approach firstly requires a traditional CFA structure, which allocates the primary indicators/items to their primary hypothesized latent factors. Factors are named on the left side of “by” and items are following on the right side. All non-primary items (i.e. crossloadings) are followed by ~ 0 to indicate approximate to 0 loadings to be calculated. For scaling purposes, the final item of each factor is followed by (*1) |
Step 2 | Testing the ESEM model | # The OUTPUT: command is followed by standardized; and stdyx; to request standardised outcomes for categorical covariates. tech 4; option is used to request estimated means, covariances, and correlations for the latent variables in the model. Finally mod (10); indicates the extraction of modification indices when the modification index for a parameter is greater than or equal to 10 OUTPUT: standardized;stdyx; tech4; mod(10); | This step produces the model results calculations |