Modelo de créditos parciais
Ricardo Primi
6 de out de 2019
Bibliotecas
library(tidyverse)
library(sjmisc)
library(TAM)
library(psych)
library(readxl)
library(knitr)
#library(xlsx)
source("http://www.labape.com.br/rprimi/R/utils_construct_maps.R")Abre banco
# Abre direto da internet (con é um endereço do arquivo)
con<-url("http://www.labape.com.br/rprimi/TRI/2019_slides_exerc/senna_v1_54_ex3.RDS")
sennav1 <- readRDS(con)
con<-url("http://www.labape.com.br/rprimi/TRI/2019_slides_exerc/senna_v1_54_ex3_dic.RDS")
dic <- readRDS(con)
# abre fazendo download do arquivo xlsx e salvando como temp.xlsx no diretório local e
# lendo com a função read_excelpath <- "http://www.labape.com.br/rprimi/TRI/2019_slides_exerc/senna_v1_54_ex3_itemdic.xlsx"
download.file(url = path, destfile = "temp.xlsx", mode="wb")
dic <- read_excel("temp.xlsx") Examina base e prepara variáveis
dic_c <- dic %>% filter(factor == "C")
vars_c <- dic_c$coditem
labels_c <- dic_c$text sennav1 %>% select(dic_c$coditem) %>% map(frq)Explorando os itens negativos, a relação entre itens semanticamente opostos
sennav1 %>% ggplot(aes(x=i83.C.Conc.11)) +
geom_histogram(
aes(y = (..count..)/sum(..count..), group = 1),
binwidth = 1,
color = "white",
fill = "blue",
alpha = 1/2
) +
theme_minimal() +
ggtitle(labels_c[7])
sennav1 %>% ggplot(aes(x=i46.C.Conc.00)) +
geom_histogram(
aes(y = (..count..)/sum(..count..), group = 1),
binwidth = 1,
color = "white",
fill = "red",
alpha = 1/2
) + theme_minimal() +
ggtitle(labels_c[3])
sennav1 %>% ggplot(aes(x=i36.C.Conc.00)) +
geom_histogram(
aes(y = (..count..)/sum(..count..), group = 1),
binwidth = 1,
color = "white",
fill = "red",
alpha = 1/2
) + theme_minimal() +
ggtitle(labels_c[2])
sennav1 %>% select(i36.C.Conc.00, i83.C.Conc.11) %>%
group_by(i36.C.Conc.00, i83.C.Conc.11) %>%
count %>%
na.omit %>%
ggplot(aes(y = i36.C.Conc.00, x = i83.C.Conc.11, fill = n)) +
geom_tile(color = "white") +
coord_fixed() +
geom_text(aes(label = n), size=4, color = "white") +
theme(legend.position="none") +
theme_minimal()
Análise psicométrica classica
sennav1 %>% select(dic_c$coditem) %>%
mutate_if(dic_c$pole == 0, funs(6 -.)) %>% # inverte negativos
na.omit %>%
alpha()##
## Reliability analysis
## Call: alpha(x = .)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.83 0.83 0.83 0.36 5 0.0026 3.4 0.73 0.37
##
## lower alpha upper 95% confidence boundaries
## 0.82 0.83 0.83
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r
## i51.C.Ord.00 0.84 0.84 0.84 0.40 5.3 0.0026 0.013
## i36.C.Conc.00 0.83 0.84 0.83 0.39 5.1 0.0026 0.015
## i46.C.Conc.00 0.81 0.82 0.81 0.36 4.4 0.0029 0.020
## i06.C.Ord.10 0.80 0.80 0.80 0.34 4.1 0.0031 0.015
## i11.C.Ord.10 0.81 0.81 0.81 0.35 4.3 0.0030 0.017
## i77.C.SD.11 0.80 0.80 0.80 0.34 4.0 0.0032 0.015
## i83.C.Conc.11 0.80 0.80 0.80 0.34 4.0 0.0031 0.016
## i89.C.Achv.11 0.81 0.81 0.81 0.35 4.4 0.0030 0.017
## i91.C.SD.11 0.81 0.81 0.81 0.35 4.3 0.0030 0.017
## med.r
## i51.C.Ord.00 0.43
## i36.C.Conc.00 0.42
## i46.C.Conc.00 0.41
## i06.C.Ord.10 0.32
## i11.C.Ord.10 0.34
## i77.C.SD.11 0.32
## i83.C.Conc.11 0.32
## i89.C.Achv.11 0.34
## i91.C.SD.11 0.34
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## i51.C.Ord.00 9383 0.46 0.46 0.34 0.31 3.7 1.12
## i36.C.Conc.00 9383 0.52 0.50 0.40 0.36 3.2 1.24
## i46.C.Conc.00 9383 0.67 0.65 0.60 0.54 3.7 1.21
## i06.C.Ord.10 9383 0.72 0.73 0.70 0.63 3.3 1.10
## i11.C.Ord.10 9383 0.69 0.69 0.64 0.57 3.1 1.16
## i77.C.SD.11 9383 0.75 0.75 0.73 0.66 3.3 1.13
## i83.C.Conc.11 9383 0.74 0.75 0.72 0.66 3.6 1.02
## i89.C.Achv.11 9383 0.64 0.66 0.60 0.54 3.4 0.97
## i91.C.SD.11 9383 0.69 0.69 0.64 0.58 3.3 1.14
##
## Non missing response frequency for each item
## 1 2 3 4 5 miss
## i51.C.Ord.00 0.05 0.09 0.26 0.31 0.28 0
## i36.C.Conc.00 0.12 0.17 0.28 0.25 0.17 0
## i46.C.Conc.00 0.07 0.10 0.22 0.29 0.31 0
## i06.C.Ord.10 0.05 0.17 0.36 0.25 0.16 0
## i11.C.Ord.10 0.08 0.24 0.33 0.20 0.15 0
## i77.C.SD.11 0.06 0.17 0.34 0.25 0.18 0
## i83.C.Conc.11 0.03 0.11 0.31 0.34 0.21 0
## i89.C.Achv.11 0.03 0.12 0.40 0.31 0.15 0
## i91.C.SD.11 0.07 0.19 0.30 0.29 0.15 0 sennav1 %>% select(x=i46.C.Conc.00, i36.C.Conc.00, i83.C.Conc.11) %>%
corr.test %>% .$r %>% kable(digits=2)
r <- cor(sennav1[ , c(vars_c, "acq_avr")],use="pair")
library(d3heatmap)
d3heatmap(r, xaxis_font_size = 10, yaxis_font_size = 10)
r <- partial.r(data = r, x = 1:9, y = 10)
d3heatmap(r, xaxis_font_size = 10, yaxis_font_size = 10)Calibra modelo de créditos parciais para o fator C (usando jml)
pc_C_jml <- sennav1 %>% select(dic_c$coditem) %>%
mutate_if(dic_c$pole == 0, funs(6 -.)) %>% # inverte negativos
na.omit %>% # elimina missings
mutate_all(funs(. -1)) %>% # tranforma escores 1-5 para 0-4
tam.jml(ndim = 1)## ....................................................
## Iteration 1 2019-10-29 18:40:06
## MLE/WLE estimation |-----
## Item parameter estimation |----------
## Deviance= 201832.6629
## Maximum MLE/WLE change: 1.572637
## Maximum item parameter change: 0.763102
## ....................................................
## Iteration 2 2019-10-29 18:40:06
## MLE/WLE estimation |----
## Item parameter estimation |----------
## Deviance= 201131.6336 | Deviance change: 701.0294
## Maximum MLE/WLE change: 0.521453
## Maximum item parameter change: 0.247805
## ....................................................
## Iteration 3 2019-10-29 18:40:06
## MLE/WLE estimation |----
## Item parameter estimation |----------
## Deviance= 200991.454 | Deviance change: 140.1796
## Maximum MLE/WLE change: 0.196873
## Maximum item parameter change: 0.094627
## ....................................................
## Iteration 4 2019-10-29 18:40:06
## MLE/WLE estimation |---
## Item parameter estimation |----------
## Deviance= 200963.0817 | Deviance change: 28.3723
## Maximum MLE/WLE change: 0.072839
## Maximum item parameter change: 0.042746
## ....................................................
## Iteration 5 2019-10-29 18:40:06
## MLE/WLE estimation |---
## Item parameter estimation |---------
## Deviance= 200957.8889 | Deviance change: 5.1928
## Maximum MLE/WLE change: 0.03185
## Maximum item parameter change: 0.019746
## ....................................................
## Iteration 6 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |--------
## Deviance= 200957.1615 | Deviance change: 0.7274
## Maximum MLE/WLE change: 0.014334
## Maximum item parameter change: 0.009091
## ....................................................
## Iteration 7 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |-------
## Deviance= 200957.1689 | Deviance change: -0.0074
## Maximum MLE/WLE change: 0.006555
## Maximum item parameter change: 0.004192
## ....................................................
## Iteration 8 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |------
## Deviance= 200957.2451 | Deviance change: -0.0762
## Maximum MLE/WLE change: 0.003004
## Maximum item parameter change: 0.00198
## ....................................................
## Iteration 9 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |----
## Deviance= 200957.2949 | Deviance change: -0.0498
## Maximum MLE/WLE change: 0.001378
## Maximum item parameter change: 0.000847
## ....................................................
## Iteration 10 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |---
## Deviance= 200957.3207 | Deviance change: -0.0258
## Maximum MLE/WLE change: 0.000658
## Maximum item parameter change: 0.000363
## ....................................................
## Iteration 11 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |--
## Deviance= 200957.3332 | Deviance change: -0.0125
## Maximum MLE/WLE change: 0.000269
## Maximum item parameter change: 0.000194
## ....................................................
## Iteration 12 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |-
## Deviance= 200957.3377 | Deviance change: -0.0045
## Maximum MLE/WLE change: 0.000131
## Maximum item parameter change: 8.9e-05
## ....................................................
## Iteration 13 2019-10-29 18:40:06
## MLE/WLE estimation |--
## Item parameter estimation |-
## Deviance= 200957.3408 | Deviance change: -0.0031
## Maximum MLE/WLE change: 7.2e-05
## Maximum item parameter change: 7e-05
##
## MLE/WLE estimation |------
## ....................................................
##
## Start: 2019-10-29 18:40:06
## End: 2019-10-29 18:40:06
## Time difference of 0.4177911 secs summary(pc_C_jml)## ------------------------------------------------------------
## TAM 3.1-45 (2019-03-18 16:53:26)
## R version 3.5.1 (2018-07-02) x86_64, darwin15.6.0 | nodename=MacBookPro2018.local | login=rprimi
##
## Start of Analysis: 2019-10-29 18:40:06
## End of Analysis: 2019-10-29 18:40:06
## Time difference of 0.4177911 secs
## Computation time: 0.4177911
##
## Joint Maximum Likelihood Estimation in TAM
##
## IRT Model
## Call:
## tam.jml(resp = ., ndim = 1)
##
## ------------------------------------------------------------
## Number of iterations= 13
##
## Deviance= 200957.3 | Log Likelihood= -100478.7
## Number of persons= 9383
## Number of items= 9
##
## Item Parameters Xsi
## xsi.label xsi.index xsi se.xsi
## 1 i51.C.Ord.00_Cat1 1 -1.577 0.053
## 2 i51.C.Ord.00_Cat2 2 -1.533 0.034
## 3 i51.C.Ord.00_Cat3 3 -0.345 0.025
## 4 i51.C.Ord.00_Cat4 4 0.481 0.027
## 5 i36.C.Conc.00_Cat1 5 -1.112 0.036
## 6 i36.C.Conc.00_Cat2 6 -0.796 0.027
## 7 i36.C.Conc.00_Cat3 7 0.156 0.026
## 8 i36.C.Conc.00_Cat4 8 1.035 0.033
## 9 i46.C.Conc.00_Cat1 9 -1.327 0.046
## 10 i46.C.Conc.00_Cat2 10 -1.258 0.033
## 11 i46.C.Conc.00_Cat3 11 -0.419 0.026
## 12 i46.C.Conc.00_Cat4 12 0.307 0.027
## 13 i06.C.Ord.10_Cat1 13 -1.919 0.050
## 14 i06.C.Ord.10_Cat2 14 -1.108 0.028
## 15 i06.C.Ord.10_Cat3 15 0.327 0.025
## 16 i06.C.Ord.10_Cat4 16 1.037 0.033
## 17 i11.C.Ord.10_Cat1 17 -1.716 0.041
## 18 i11.C.Ord.10_Cat2 18 -0.617 0.026
## 19 i11.C.Ord.10_Cat3 19 0.561 0.027
## 20 i11.C.Ord.10_Cat4 20 1.028 0.036
## 21 i77.C.SD.11_Cat1 21 -1.919 0.050
## 22 i77.C.SD.11_Cat2 22 -1.042 0.028
## 23 i77.C.SD.11_Cat3 23 0.278 0.025
## 24 i77.C.SD.11_Cat4 24 0.882 0.032
## 25 i83.C.Conc.11_Cat1 25 -2.379 0.070
## 26 i83.C.Conc.11_Cat2 26 -1.509 0.034
## 27 i83.C.Conc.11_Cat3 27 -0.190 0.024
## 28 i83.C.Conc.11_Cat4 28 0.905 0.030
## 29 i89.C.Achv.11_Cat1 29 -2.370 0.069
## 30 i89.C.Achv.11_Cat2 30 -1.592 0.033
## 31 i89.C.Achv.11_Cat3 31 0.188 0.024
## 32 i89.C.Achv.11_Cat4 32 1.275 0.034
## 33 i91.C.SD.11_Cat1 33 -1.736 0.045
## 34 i91.C.SD.11_Cat2 34 -0.839 0.028
## 35 i91.C.SD.11_Cat3 35 0.062 0.025
## 36 i91.C.SD.11_Cat4 36 1.220 0.034 fit <- tam.jml.fit(pc_C_jml)
fit$fit.item## item outfitItem outfitItem_t infitItem infitItem_t
## i51.C.Ord.00 i51.C.Ord.00 1.4652471 23.226296 1.3569468 22.235892
## i36.C.Conc.00 i36.C.Conc.00 1.4530288 25.126186 1.3528156 23.259053
## i46.C.Conc.00 i46.C.Conc.00 0.9434446 -3.092075 0.9370100 -4.470045
## i06.C.Ord.10 i06.C.Ord.10 0.7619128 -16.759013 0.7669273 -18.238869
## i11.C.Ord.10 i11.C.Ord.10 0.9001655 -6.614164 0.8939265 -7.970278
## i77.C.SD.11 i77.C.SD.11 0.7181314 -19.787725 0.7264913 -21.884254
## i83.C.Conc.11 i83.C.Conc.11 0.6788896 -22.988429 0.6925441 -24.425037
## i89.C.Achv.11 i89.C.Achv.11 0.8738759 -8.626175 0.8704372 -9.473562
## i91.C.SD.11 i91.C.SD.11 0.8831716 -8.029612 0.8727974 -9.670340 plot(pc_C_jml, ngroups = 24)
## ....................................................
## Plots exported in png format into folder:
## /Users/rprimi/Dropbox (Personal)/TRI/2019_slides_exerc/Plots plot(pc_C_jml, type = "items")
## ....................................................
## Plots exported in png format into folder:
## /Users/rprimi/Dropbox (Personal)/TRI/2019_slides_exerc/PlotsMapa de construto
pc_C <- sennav1 %>% select(dic_c$coditem) %>%
mutate_if(dic_c$pole == 0, funs(6 -.)) %>% # inverte negativos
mutate_all(funs(. -1)) %>% # tranforma escores 1-5 para 0-4
tam.mml(irtmodel = "PCM" )## ....................................................
## Processing Data 2019-10-29 18:40:09
## * Response Data: 11249 Persons and 9 Items
## * Numerical integration with 21 nodes
## * Created Design Matrices ( 2019-10-29 18:40:09 )
## * Calculated Sufficient Statistics ( 2019-10-29 18:40:10 )
## ....................................................
## Iteration 1 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 267613.0422
## Maximum item intercept parameter change: 0.65952
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.155948
## ....................................................
## Iteration 2 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 263472.3794 | Absolute change: 4140.663 | Relative change: 0.01571574
## Maximum item intercept parameter change: 0.494493
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.166203
## ....................................................
## Iteration 3 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262593.7818 | Absolute change: 878.5976 | Relative change: 0.00334584
## Maximum item intercept parameter change: 0.130315
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.121802
## ....................................................
## Iteration 4 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262305.3891 | Absolute change: 288.3927 | Relative change: 0.00109945
## Maximum item intercept parameter change: 0.069541
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.070951
## ....................................................
## Iteration 5 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262188.9176 | Absolute change: 116.4715 | Relative change: 0.00044423
## Maximum item intercept parameter change: 0.044501
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.043275
## ....................................................
## Iteration 6 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262138.9315 | Absolute change: 49.986 | Relative change: 0.00019069
## Maximum item intercept parameter change: 0.026189
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.02709
## ....................................................
## Iteration 7 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262116.3339 | Absolute change: 22.5976 | Relative change: 8.621e-05
## Maximum item intercept parameter change: 0.020116
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.017424
## ....................................................
## Iteration 8 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262105.6179 | Absolute change: 10.716 | Relative change: 4.088e-05
## Maximum item intercept parameter change: 0.012089
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.011353
## ....................................................
## Iteration 9 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262100.4057 | Absolute change: 5.2122 | Relative change: 1.989e-05
## Maximum item intercept parameter change: 0.008415
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.007294
## ....................................................
## Iteration 10 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262097.5434 | Absolute change: 2.8623 | Relative change: 1.092e-05
## Maximum item intercept parameter change: 0.006195
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.004862
## ....................................................
## Iteration 11 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262095.8528 | Absolute change: 1.6906 | Relative change: 6.45e-06
## Maximum item intercept parameter change: 0.004662
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.003303
## ....................................................
## Iteration 12 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262094.7783 | Absolute change: 1.0746 | Relative change: 4.1e-06
## Maximum item intercept parameter change: 0.003581
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.002319
## ....................................................
## Iteration 13 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262094.0693 | Absolute change: 0.7089 | Relative change: 2.7e-06
## Maximum item intercept parameter change: 0.002862
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.001686
## ....................................................
## Iteration 14 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262093.6007 | Absolute change: 0.4686 | Relative change: 1.79e-06
## Maximum item intercept parameter change: 0.002256
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.00127
## ....................................................
## Iteration 15 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262093.2826 | Absolute change: 0.3182 | Relative change: 1.21e-06
## Maximum item intercept parameter change: 0.00185
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000948
## ....................................................
## Iteration 16 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262093.0584 | Absolute change: 0.2241 | Relative change: 8.6e-07
## Maximum item intercept parameter change: 0.001514
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000721
## ....................................................
## Iteration 17 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.9002 | Absolute change: 0.1582 | Relative change: 6e-07
## Maximum item intercept parameter change: 0.00125
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000564
## ....................................................
## Iteration 18 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.7864 | Absolute change: 0.1137 | Relative change: 4.3e-07
## Maximum item intercept parameter change: 0.001032
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000431
## ....................................................
## Iteration 19 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.7038 | Absolute change: 0.0827 | Relative change: 3.2e-07
## Maximum item intercept parameter change: 0.00086
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000332
## ....................................................
## Iteration 20 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.6433 | Absolute change: 0.0605 | Relative change: 2.3e-07
## Maximum item intercept parameter change: 0.000722
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000261
## ....................................................
## Iteration 21 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5991 | Absolute change: 0.0442 | Relative change: 1.7e-07
## Maximum item intercept parameter change: 0.00059
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.00021
## ....................................................
## Iteration 22 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5665 | Absolute change: 0.0325 | Relative change: 1.2e-07
## Maximum item intercept parameter change: 0.000508
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000167
## ....................................................
## Iteration 23 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5424 | Absolute change: 0.0241 | Relative change: 9e-08
## Maximum item intercept parameter change: 0.000436
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.000134
## ....................................................
## Iteration 24 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5245 | Absolute change: 0.0179 | Relative change: 7e-08
## Maximum item intercept parameter change: 0.000374
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 0.00011
## ....................................................
## Iteration 25 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5111 | Absolute change: 0.0133 | Relative change: 5e-08
## Maximum item intercept parameter change: 0.00032
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 9e-05
## ....................................................
## Iteration 26 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.5012 | Absolute change: 0.0099 | Relative change: 4e-08
## Maximum item intercept parameter change: 0.000275
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 7.5e-05
## ....................................................
## Iteration 27 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.4938 | Absolute change: 0.0074 | Relative change: 3e-08
## Maximum item intercept parameter change: 0.000236
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 6.3e-05
## ....................................................
## Iteration 28 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.4884 | Absolute change: 0.0054 | Relative change: 2e-08
## Maximum item intercept parameter change: 0.000203
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 5.5e-05
## ....................................................
## Iteration 29 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.4843 | Absolute change: 0.0041 | Relative change: 2e-08
## Maximum item intercept parameter change: 0.000174
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 4.6e-05
## ....................................................
## Iteration 30 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.4813 | Absolute change: 0.003 | Relative change: 1e-08
## Maximum item intercept parameter change: 0.000149
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 3.9e-05
## ....................................................
## Iteration 31 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.479 | Absolute change: 0.0023 | Relative change: 1e-08
## Maximum item intercept parameter change: 0.000129
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 3.3e-05
## ....................................................
## Iteration 32 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.4773 | Absolute change: 0.0017 | Relative change: 1e-08
## Maximum item intercept parameter change: 0.000111
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 2.8e-05
## ....................................................
## Iteration 33 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.476 | Absolute change: 0.0013 | Relative change: 0
## Maximum item intercept parameter change: 9.6e-05
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 2.4e-05
## ....................................................
## Iteration 34 2019-10-29 18:40:10
## E Step
## M Step Intercepts |----
## Deviance = 262092.475 | Absolute change: 0.001 | Relative change: 0
## Maximum item intercept parameter change: 8.3e-05
## Maximum item slope parameter change: 0
## Maximum regression parameter change: 0
## Maximum variance parameter change: 2e-05
## ....................................................
## Item Parameters
## xsi.index xsi.label est
## 1 1 i51.C.Ord.00_Cat1 -1.4658
## 2 2 i51.C.Ord.00_Cat2 -1.5354
## 3 3 i51.C.Ord.00_Cat3 -0.2853
## 4 4 i51.C.Ord.00_Cat4 0.4178
## 5 5 i36.C.Conc.00_Cat1 -1.0007
## 6 6 i36.C.Conc.00_Cat2 -0.7326
## 7 7 i36.C.Conc.00_Cat3 0.1900
## 8 8 i36.C.Conc.00_Cat4 0.9531
## 9 9 i46.C.Conc.00_Cat1 -1.2001
## 10 10 i46.C.Conc.00_Cat2 -1.1976
## 11 11 i46.C.Conc.00_Cat3 -0.3775
## 12 12 i46.C.Conc.00_Cat4 0.2629
## 13 13 i06.C.Ord.10_Cat1 -1.8128
## 14 14 i06.C.Ord.10_Cat2 -1.0699
## 15 15 i06.C.Ord.10_Cat3 0.3958
## 16 16 i06.C.Ord.10_Cat4 0.9461
## 17 17 i11.C.Ord.10_Cat1 -1.6292
## 18 18 i11.C.Ord.10_Cat2 -0.5619
## 19 19 i11.C.Ord.10_Cat3 0.6342
## 20 20 i11.C.Ord.10_Cat4 0.8921
## 21 21 i77.C.SD.11_Cat1 -1.8057
## 22 22 i77.C.SD.11_Cat2 -1.0159
## 23 23 i77.C.SD.11_Cat3 0.3515
## 24 24 i77.C.SD.11_Cat4 0.7918
## 25 25 i83.C.Conc.11_Cat1 -2.2297
## 26 26 i83.C.Conc.11_Cat2 -1.4733
## 27 27 i83.C.Conc.11_Cat3 -0.1257
## 28 28 i83.C.Conc.11_Cat4 0.8509
## 29 29 i89.C.Achv.11_Cat1 -2.2233
## 30 30 i89.C.Achv.11_Cat2 -1.5677
## 31 31 i89.C.Achv.11_Cat3 0.2809
## 32 32 i89.C.Achv.11_Cat4 1.2200
## 33 33 i91.C.SD.11_Cat1 -1.6386
## 34 34 i91.C.SD.11_Cat2 -0.7918
## 35 35 i91.C.SD.11_Cat3 0.0905
## 36 36 i91.C.SD.11_Cat4 1.1476
## ...................................
## Regression Coefficients
## [,1]
## [1,] 0
##
## Variance:
## [,1]
## [1,] 0.6728
##
##
## EAP Reliability:
## [1] 0.804
##
## -----------------------------
## Start: 2019-10-29 18:40:09
## End: 2019-10-29 18:40:10
## Time difference of 0.7537041 secs summary(pc_C)## ------------------------------------------------------------
## TAM 3.1-45 (2019-03-18 16:53:26)
## R version 3.5.1 (2018-07-02) x86_64, darwin15.6.0 | nodename=MacBookPro2018.local | login=rprimi
##
## Date of Analysis: 2019-10-29 18:40:10
## Time difference of 0.7537041 secs
## Computation time: 0.7537041
##
## Multidimensional Item Response Model in TAM
##
## IRT Model: PCM
## Call:
## tam.mml(resp = ., irtmodel = "PCM")
##
## ------------------------------------------------------------
## Number of iterations = 34
## Numeric integration with 21 integration points
##
## Deviance = 262092.5
## Log likelihood = -131046.2
## Number of persons = 11249
## Number of persons used = 11029
## Number of items = 9
## Number of estimated parameters = 37
## Item threshold parameters = 36
## Item slope parameters = 0
## Regression parameters = 0
## Variance/covariance parameters = 1
##
## AIC = 262166 | penalty = 74 | AIC=-2*LL + 2*p
## AIC3 = 262203 | penalty = 111 | AIC3=-2*LL + 3*p
## BIC = 262437 | penalty = 344.41 | BIC=-2*LL + log(n)*p
## aBIC = 262319 | penalty = 226.81 | aBIC=-2*LL + log((n-2)/24)*p (adjusted BIC)
## CAIC = 262474 | penalty = 381.41 | CAIC=-2*LL + [log(n)+1]*p (consistent AIC)
## AICc = 262167 | penalty = 74.26 | AICc=-2*LL + 2*p + 2*p*(p+1)/(n-p-1) (bias corrected AIC)
##
## ------------------------------------------------------------
## EAP Reliability
## [1] 0.804
## ------------------------------------------------------------
## Covariances and Variances
## [,1]
## [1,] 0.673
## ------------------------------------------------------------
## Correlations and Standard Deviations (in the diagonal)
## [,1]
## [1,] 0.82
## ------------------------------------------------------------
## Regression Coefficients
## [,1]
## [1,] 0
## ------------------------------------------------------------
## Item Parameters -A*Xsi
## item N M xsi.item AXsi_.Cat1 AXsi_.Cat2
## i51.C.Ord.00 i51.C.Ord.00 10685 2.685 -0.717 -1.466 -3.001
## i36.C.Conc.00 i36.C.Conc.00 10605 2.168 -0.148 -1.001 -1.733
## i46.C.Conc.00 i46.C.Conc.00 10717 2.657 -0.628 -1.200 -2.398
## i06.C.Ord.10 i06.C.Ord.10 10629 2.295 -0.385 -1.813 -2.883
## i11.C.Ord.10 i11.C.Ord.10 10671 2.081 -0.166 -1.629 -2.191
## i77.C.SD.11 i77.C.SD.11 10731 2.324 -0.420 -1.806 -2.822
## i83.C.Conc.11 i83.C.Conc.11 10689 2.583 -0.745 -2.230 -3.703
## i89.C.Achv.11 i89.C.Achv.11 10708 2.412 -0.573 -2.223 -3.791
## i91.C.SD.11 i91.C.SD.11 10691 2.267 -0.298 -1.639 -2.431
## AXsi_.Cat3 AXsi_.Cat4 B.Cat1.Dim1 B.Cat2.Dim1 B.Cat3.Dim1
## i51.C.Ord.00 -3.287 -2.869 1 2 3
## i36.C.Conc.00 -1.543 -0.590 1 2 3
## i46.C.Conc.00 -2.775 -2.513 1 2 3
## i06.C.Ord.10 -2.487 -1.541 1 2 3
## i11.C.Ord.10 -1.557 -0.665 1 2 3
## i77.C.SD.11 -2.470 -1.679 1 2 3
## i83.C.Conc.11 -3.829 -2.978 1 2 3
## i89.C.Achv.11 -3.510 -2.290 1 2 3
## i91.C.SD.11 -2.340 -1.193 1 2 3
## B.Cat4.Dim1
## i51.C.Ord.00 4
## i36.C.Conc.00 4
## i46.C.Conc.00 4
## i06.C.Ord.10 4
## i11.C.Ord.10 4
## i77.C.SD.11 4
## i83.C.Conc.11 4
## i89.C.Achv.11 4
## i91.C.SD.11 4
##
## Item Parameters Xsi
## xsi se.xsi
## i51.C.Ord.00_Cat1 -1.466 0.048
## i51.C.Ord.00_Cat2 -1.535 0.031
## i51.C.Ord.00_Cat3 -0.285 0.023
## i51.C.Ord.00_Cat4 0.418 0.024
## i36.C.Conc.00_Cat1 -1.001 0.033
## i36.C.Conc.00_Cat2 -0.733 0.025
## i36.C.Conc.00_Cat3 0.190 0.023
## i36.C.Conc.00_Cat4 0.953 0.029
## i46.C.Conc.00_Cat1 -1.200 0.041
## i46.C.Conc.00_Cat2 -1.198 0.029
## i46.C.Conc.00_Cat3 -0.377 0.023
## i46.C.Conc.00_Cat4 0.263 0.024
## i06.C.Ord.10_Cat1 -1.813 0.044
## i06.C.Ord.10_Cat2 -1.070 0.026
## i06.C.Ord.10_Cat3 0.396 0.023
## i06.C.Ord.10_Cat4 0.946 0.029
## i11.C.Ord.10_Cat1 -1.629 0.037
## i11.C.Ord.10_Cat2 -0.562 0.024
## i11.C.Ord.10_Cat3 0.634 0.024
## i11.C.Ord.10_Cat4 0.892 0.031
## i77.C.SD.11_Cat1 -1.806 0.044
## i77.C.SD.11_Cat2 -1.016 0.026
## i77.C.SD.11_Cat3 0.352 0.023
## i77.C.SD.11_Cat4 0.792 0.028
## i83.C.Conc.11_Cat1 -2.230 0.060
## i83.C.Conc.11_Cat2 -1.473 0.030
## i83.C.Conc.11_Cat3 -0.126 0.022
## i83.C.Conc.11_Cat4 0.851 0.026
## i89.C.Achv.11_Cat1 -2.223 0.059
## i89.C.Achv.11_Cat2 -1.568 0.029
## i89.C.Achv.11_Cat3 0.281 0.022
## i89.C.Achv.11_Cat4 1.220 0.030
## i91.C.SD.11_Cat1 -1.639 0.040
## i91.C.SD.11_Cat2 -0.792 0.025
## i91.C.SD.11_Cat3 0.091 0.023
## i91.C.SD.11_Cat4 1.148 0.030
##
## Item Parameters in IRT parameterization
## item alpha beta tau.Cat1 tau.Cat2 tau.Cat3 tau.Cat4
## 1 i51.C.Ord.00 1 -0.717 -0.749 -0.818 0.432 1.135
## 2 i36.C.Conc.00 1 -0.148 -0.853 -0.585 0.338 1.101
## 3 i46.C.Conc.00 1 -0.628 -0.572 -0.570 0.251 0.891
## 4 i06.C.Ord.10 1 -0.385 -1.428 -0.685 0.781 1.331
## 5 i11.C.Ord.10 1 -0.166 -1.463 -0.396 0.800 1.058
## 6 i77.C.SD.11 1 -0.420 -1.386 -0.596 0.771 1.211
## 7 i83.C.Conc.11 1 -0.745 -1.485 -0.729 0.619 1.595
## 8 i89.C.Achv.11 1 -0.573 -1.651 -0.995 0.853 1.793
## 9 i91.C.SD.11 1 -0.298 -1.341 -0.494 0.389 1.446 IRT.WrightMap(pc_C, label.items = labels_c)
tam.threshold(pc_C)## Cat1 Cat2 Cat3 Cat4
## i51.C.Ord.00 -2.017548 -1.2876892 -0.3355408 0.7506409
## i36.C.Conc.00 -1.460541 -0.6319885 0.2101135 1.2794495
## i46.C.Conc.00 -1.752960 -1.0633850 -0.3408508 0.6246643
## i06.C.Ord.10 -2.131073 -0.9690857 0.2527771 1.3103943
## i11.C.Ord.10 -1.883514 -0.5754089 0.4679260 1.3428040
## i77.C.SD.11 -2.115326 -0.9430847 0.2005920 1.1866150
## i83.C.Conc.11 -2.546906 -1.3876648 -0.1674500 1.1206970
## i89.C.Achv.11 -2.555878 -1.3905945 0.1604919 1.4918518
## i91.C.SD.11 -1.945221 -0.8103333 0.1509705 1.4089050Mapa de construto: tentativa 2
library(WrightMap)
thurst_gam <- as.data.frame(tam.threshold(pc_C))
thurst_gam <- row_means( thurst_gam, n = 4)
thurst_gam$item_label <- dic_c$text
thurst_gam <- thurst_gam %>% arrange(rowmeans)
names(thurst_gam)[1:4] <-c("s2", "s3", "s4", "s5" )
thetas <- pc_C$person$EAP
summary(thetas)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -2.625638 -0.503487 0.000000 -0.001252 0.455481 2.029737# Mapa mais elaborado
# Cores dos thresholds
library(RColorBrewer)
cores <- rep(brewer.pal(4, "Set1"), 9)
threshold_col <- matrix(cores, byrow = TRUE, ncol = 4)
# Mapa de construto
wrightMap(thetas, thurst_gam[ , 1:4],
main.title = "",
item.prop = 0.75, # Proporçao espaço do item/theta
show.thr.lab= TRUE, # Nao mostra label dos thresholds
thr.sym.col.fg = threshold_col, # Colori os thresholds
thr.sym.col.bg = threshold_col,
thr.sym.cex = .8, # tamanho dos simbolos do thresshold
axis.items="", # elimina label do eixo x
label.items.srt=25, # ajusta item label para vertical
label.items.cex = .45, # tamanho da fonte dos item
label.items = thurst_gam$item_label, # label dos itens
return.thresholds = FALSE, cutpoints = c(-.50, 0, .45)
)
Mapa de construto: tentativa 2
ver: https://cran.r-project.org/web/packages/eRm/vignettes/eRm.pdf
library(eRm)
res.pcm <- sennav1 %>% select(dic_c$coditem) %>%
mutate_if(dic_c$pole == 0, funs(6 -.)) %>% # inverte negativos
mutate_all(funs(. -1)) %>%
na.omit %>%
PCM
plotPImap(res.pcm , sorted = TRUE )
Mapa de construto: tentativa 3
person_item_map_v2(
item_tresh = tam.threshold(pc_C),
coditem = dic_c$coditem,
item_text = dic_c$text,
pole = dic_c$pole,
theta = pc_C$person$EAP
)
describe_likert5_items(
data = sennav1[ , dic_c$coditem],
item_tresh = tam.threshold(pc_C),
coditem = dic_c$coditem,
item_text = dic_c$text,
pole = dic_c$pole
)
Mapa de construto: tentativa 4. Agora Sim!
person_item_map_v3(
item_tresh = tam.threshold(pc_C),
coditem = dic_c$coditem,
item_text = dic_c$text,
pole = dic_c$pole,
theta = pc_C$person$EAP,
size_categ_label = 3,
size_bar = 4.5,
item_text_max = 32
)
Tente você
- Escolha uma escala e faça as mesmas análises
- Crie uma interpretação substancial para a escala