john_Analysis
Sebastian Robledo
3/9/2022
ArtÃculo 4
Cleaning data
Citation network
ArtÃculo 3 - LM
Tabla
Estimate | Standard Error | t value | Pr(>|t|) | ||
(Intercept) | 0.183 | 0.120 | 1.530 | 0.1264 | |
EE | 0.512 | 0.029 | 17.524 | 0.0000 | *** |
IC | 0.431 | 0.027 | 15.856 | 0.0000 | *** |
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1 | |||||
Residual standard error: 0.5674 on 997 degrees of freedom | |||||
Multiple R-squared: 0.3561, Adjusted R-squared: 0.3548 | |||||
F-statistic: 275.6 on 997 and 2 DF, p-value: 0.0000 | |||||
Análisis
## We fitted a linear model (estimated using OLS) to predict RE with EE and IC (formula: RE ~ EE + IC). The model explains a statistically significant and substantial proportion of variance (R2 = 0.36, F(2, 997) = 275.64, p < .001, adj. R2 = 0.35). The model's intercept, corresponding to EE = 0 and IC = 0, is at 0.18 (95% CI [-0.05, 0.42], t(997) = 1.53, p = 0.126). Within this model:
##
## - The effect of EE is statistically significant and positive (beta = 0.51, 95% CI [0.46, 0.57], t(997) = 17.52, p < .001; Std. beta = 0.45, 95% CI [0.40, 0.50])
## - The effect of IC is statistically significant and positive (beta = 0.43, 95% CI [0.38, 0.48], t(997) = 15.86, p < .001; Std. beta = 0.40, 95% CI [0.35, 0.45])
##
## Standardized parameters were obtained by fitting the model on a standardized version of the dataset. 95% Confidence Intervals (CIs) and p-values were computed using the Wald approximation.ArtÃculo 2 - GLM
Tabla
Estimate | Standard Error | z value | Pr(>|z|) | Signif. | |
(Intercept) | 7.951 | 0.615 | 12.925 | 0.0000 | *** |
EE | -1.721 | 0.143 | -11.999 | 0.0000 | *** |
IC | -1.165 | 0.124 | -9.424 | 0.0000 | *** |
Signif. codes: 0 <= '***' < 0.001 < '**' < 0.01 < '*' < 0.05 < '.' < 0.1 < '' < 1 | |||||
| |||||
(Dispersion parameter for binomial family taken to be 1) | |||||
Null deviance: 1345 on 999 degrees of freedom | |||||
Residual deviance: 1077 on 997 degrees of freedom | |||||
Análisis
## We fitted a logistic model (estimated using ML) to predict RE with EE and IC (formula: RE ~ EE + IC). The model's explanatory power is moderate (Tjur's R2 = 0.24). The model's intercept, corresponding to EE = 0 and IC = 0, is at 7.95 (95% CI [6.78, 9.19], p < .001). Within this model:
##
## - The effect of EE is statistically significant and negative (beta = -1.72, 95% CI [-2.01, -1.45], p < .001; Std. beta = -1.06, 95% CI [-1.23, -0.89])
## - The effect of IC is statistically significant and negative (beta = -1.17, 95% CI [-1.41, -0.93], p < .001; Std. beta = -0.77, 95% CI [-0.93, -0.61])
##
## Standardized parameters were obtained by fitting the model on a standardized version of the dataset. 95% Confidence Intervals (CIs) and p-values were computed usingArtÃculo 1 - SEM
SEM Análisis
Simulación de datos
RE = Racismo Escolar
EE = Etnoeducación
IC = Interculturalidad
El Racismo Escolar es influenciada por la Etnoeducacióny por la Interculturalidad
PC ~ EE + CRP
Análisis Exploratorio
factanal(johnSimData, factors = 3, rotation = "promax")##
## Call:
## factanal(x = johnSimData, factors = 3, rotation = "promax")
##
## Uniquenesses:
## i1 i2 i3 i4 i5 i6 i7 i8 i9
## 0.598 0.462 0.635 0.507 0.547 0.517 0.403 0.333 0.392
##
## Loadings:
## Factor1 Factor2 Factor3
## i1 0.102 0.575
## i2 0.775
## i3 0.599
## i4 0.706
## i5 0.663
## i6 0.692
## i7 0.767
## i8 0.837
## i9 0.738
##
## Factor1 Factor2 Factor3
## SS loadings 1.851 1.428 1.292
## Proportion Var 0.206 0.159 0.144
## Cumulative Var 0.206 0.364 0.508
##
## Factor Correlations:
## Factor1 Factor2 Factor3
## Factor1 1.000 -0.4796 0.5274
## Factor2 -0.480 1.0000 0.0161
## Factor3 0.527 0.0161 1.0000
##
## Test of the hypothesis that 3 factors are sufficient.
## The chi square statistic is 8.41 on 12 degrees of freedom.
## The p-value is 0.752| Valor | Concepto |
|---|---|
| < 0.39 | pobre |
| .4 - .49 | Justo |
| .5 - .59 | Bueno |
| .6 - .69 | Muy bueno |
| .7 + | Excelente |
Modelo
Creamos el modelo
johnModel <-
"EE =~ i1 + i2 + i3
IC =~ i4 + i5 + i6
RE =~ i7 + i8 + i9
RE ~ EE
RE ~ IC"
john.fit <-
cfa(johnModel,
data = johnSimData)Revisamos las cargas
inspect(john.fit,
what = "std")$lambda## EE IC RE
## i1 0.659 0.000 0.000
## i2 0.696 0.000 0.000
## i3 0.607 0.000 0.000
## i4 0.000 0.694 0.000
## i5 0.000 0.680 0.000
## i6 0.000 0.695 0.000
## i7 0.000 0.000 0.774
## i8 0.000 0.000 0.806
## i9 0.000 0.000 0.788>.6
Evaluamos el modelo
summary(john.fit,
# standardized = TRUE,
fit.measures = TRUE)## lavaan 0.6-9 ended normally after 33 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 21
##
## Number of observations 1000
##
## Model Test User Model:
##
## Test statistic 19.376
## Degrees of freedom 24
## P-value (Chi-square) 0.732
##
## Model Test Baseline Model:
##
## Test statistic 2668.520
## Degrees of freedom 36
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 1.000
## Tucker-Lewis Index (TLI) 1.003
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -9416.917
## Loglikelihood unrestricted model (H1) -9407.229
##
## Akaike (AIC) 18875.835
## Bayesian (BIC) 18978.897
## Sample-size adjusted Bayesian (BIC) 18912.200
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.000
## 90 Percent confidence interval - lower 0.000
## 90 Percent confidence interval - upper 0.019
## P-value RMSEA <= 0.05 1.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.015
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## EE =~
## i1 1.000
## i2 0.961 0.067 14.399 0.000
## i3 0.811 0.059 13.744 0.000
## IC =~
## i4 1.000
## i5 1.102 0.070 15.824 0.000
## i6 0.981 0.061 15.950 0.000
## RE =~
## i7 1.000
## i8 1.191 0.050 23.949 0.000
## i9 1.245 0.053 23.580 0.000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|)
## RE ~
## EE 0.518 0.044 11.758 0.000
## IC 0.577 0.048 12.043 0.000
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## EE ~~
## IC 0.000 0.013 0.001 0.999
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .i1 0.443 0.029 15.488 0.000
## .i2 0.333 0.024 13.971 0.000
## .i3 0.382 0.022 17.198 0.000
## .i4 0.273 0.018 15.145 0.000
## .i5 0.359 0.023 15.698 0.000
## .i6 0.261 0.017 15.086 0.000
## .i7 0.219 0.014 16.027 0.000
## .i8 0.250 0.017 14.447 0.000
## .i9 0.309 0.020 15.374 0.000
## EE 0.339 0.035 9.569 0.000
## IC 0.254 0.024 10.511 0.000
## .RE 0.152 0.015 9.956 0.000| Medidas | Valor a revisar |
|---|---|
| chi-square | p < .05 |
| CFI - Comparative Fit Index | >.9 |
| TLI - Tuker-Lewis Index | >.9 |
| RMSEA | <.05 |
Tabla 1
Promedios, desviaciones estándar, confiabilidad interna y variables compuestas (N = 1000)
john_mean <-
johnSimData |>
rowwise() |>
mutate(RE = mean(i1, i2, i3),
DE = mean(i4, i5, i6),
CD = mean(i7, i8, i9)) |>
dplyr::select(RE, DE, CD) |>
ungroup() |>
summarise(across(everything(), mean))
john_sd <-
johnSimData |>
rowwise() |>
mutate(RE = mean(i1, i2, i3),
DE = mean(i4, i5, i6),
CD = mean(i7, i8, i9)) |>
dplyr::select(RE, DE, CD) |>
ungroup() |>
summarise(across(everything(), sd))
john_sd_mean <-
john_mean |>
bind_rows(john_sd) |>
transpose_df() |>
rename(Variable_compuesta = rowname,
promedio = "1",
des_est = "2") |>
mutate(promedio = round(promedio, digits = 2),
des_est = round(des_est, digits = 2))
john_cronbach_RE <-
johnSimData |>
dplyr::select(i1, i2, i3) |>
cronbach.alpha()
john_cronbach_DE <-
johnSimData |>
dplyr::select(i4, i5, i6) |>
cronbach.alpha()
john_cronbach_CD <-
johnSimData |>
dplyr::select(i7, i8, i9) |>
cronbach.alpha()
john_cronbach <-
tibble(RE = john_cronbach_RE$alpha,
DE = john_cronbach_DE$alpha,
CD = john_cronbach_CD$alpha) |>
transpose_df() |>
rename(Variable_compuesta = rowname,
Cronbach = "1") |>
right_join(john_sd_mean) |>
dplyr::select(Variable_compuesta,
Cronbach,
promedio,
des_est) |>
mutate(Cronbach = round(Cronbach,
digits = 2),
No_items = 3) |>
dplyr::select(Variable_compuesta,
No_items,
Cronbach,
promedio,
des_est
)## Joining, by = "Variable_compuesta"rm(john_mean,
john_sd,
john_sd_mean)
john_cronbach |>
gt() |>
tab_header(
title = "Tabla 1 Promedios, desviaciones estándar, confiabilidad interna y variables compuestas (N = 1000)"
)| Tabla 1 Promedios, desviaciones estándar, confiabilidad interna y variables compuestas (N = 1000) | ||||
|---|---|---|---|---|
| Variable_compuesta | No_items | Cronbach | promedio | des_est |
| RE | 3 | 0.69 | 2.90 | 0.88 |
| DE | 3 | 0.73 | 2.97 | 0.73 |
| CD | 3 | 0.83 | 2.75 | 0.74 |
EstadÃsticas a tener en cuenta
| Cronbach Alpha | Concepto |
|---|---|
| <.6 | Inaceptable - Se eliminan items |
| 6 - .64 | Indeseable |
| .65 - .69 | Aceptable minimamente |
| .7 - .79 | Respetable |
| .8 - .89 | Muy bueno |
| .9> | Demasiado buenos - eliminar items |
Tabla 2
Matrix de correlaciones
RE <- johnSimData |>
rowwise() |>
mutate(RE = mean(i1, i2, i3))
DE <- johnSimData |> rowwise() |> mutate(DE = mean(i4, i5, i6))
CD <- johnSimData |> rowwise() |> mutate(CD = mean(i7, i8, i9))
john_latent <-
tibble(RE = RE$RE,
DE = DE$DE,
CD = CD$CD)
rcorr(as.matrix(john_latent))## RE DE CD
## RE 1.00 -0.02 0.29
## DE -0.02 1.00 0.27
## CD 0.29 0.27 1.00
##
## n= 1000
##
##
## P
## RE DE CD
## RE 0.437 0.000
## DE 0.437 0.000
## CD 0.000 0.000Diagrama del modelo
semPaths(john.fit,
what = "est",
fade = FALSE,
residuals = FALSE,
edge.label.cex = 0.75)