Alberts_APA_2019
Patrick N. Beymer
11/21/2018
Loading packages
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.2.1 ──## ✔ ggplot2 3.0.0 ✔ readr 1.1.1
## ✔ tibble 1.4.2 ✔ purrr 0.2.5
## ✔ tidyr 0.8.1 ✔ stringr 1.3.1
## ✔ ggplot2 3.0.0 ✔ forcats 0.3.0## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(lme4)## Loading required package: Matrix##
## Attaching package: 'Matrix'## The following object is masked from 'package:tidyr':
##
## expandlibrary(lmerTest)##
## Attaching package: 'lmerTest'## The following object is masked from 'package:lme4':
##
## lmer## The following object is masked from 'package:stats':
##
## steplibrary(sjstats)
library(jmRtools)
library(MuMIn)
library(konfound)## Sensitivity analysis as described in Frank, Maroulis, Duong, and Kelcey (2013) and in Frank (2000).
## For more information visit https://jmichaelrosenberg.shinyapps.io/shinykonfound/.Loading data
imuscle_esm <- read_csv("iMUScLE-esm.csv")## Parsed with column specification:
## cols(
## .default = col_double(),
## TEACHID = col_integer(),
## PERIOD = col_integer(),
## MONTH = col_integer(),
## DAY = col_integer(),
## YEAR = col_integer(),
## time = col_integer(),
## timepoint = col_integer(),
## STUDID = col_integer(),
## ENJOY = col_integer(),
## group = col_integer(),
## brain = col_integer(),
## grade = col_integer(),
## ESF_counts = col_integer(),
## Uti7T = col_integer(),
## Uti7S = col_integer(),
## Uti79TS = col_integer(),
## Uti7T9B = col_integer(),
## Use_r = col_integer(),
## Uful_r = col_integer(),
## Ugroup = col_integer()
## # ... with 4 more columns
## )## See spec(...) for full column specifications.## Warning in rbind(names(probs), probs_f): number of columns of result is not
## a multiple of vector length (arg 1)## Warning: 7029 parsing failures.
## row # A tibble: 5 x 5 col row col expected actual file expected <int> <chr> <chr> <chr> <chr> actual 1 3172 Uti7T an integer NaN 'iMUScLE-esm.csv' file 2 3172 Uti7S an integer NaN 'iMUScLE-esm.csv' row 3 3173 Uti7T an integer NaN 'iMUScLE-esm.csv' col 4 3173 Uti7S an integer NaN 'iMUScLE-esm.csv' expected 5 3174 Uti7T an integer NaN 'iMUScLE-esm.csv'
## ... ................. ... ................................................. ........ ................................................. ...... ................................................. .... ................................................. ... ................................................. ... ................................................. ........ .................................................
## See problems(...) for more details.imuscle_student_survey <- read_csv("iMUScLE-student-survey.csv")## Parsed with column specification:
## cols(
## .default = col_double(),
## STUDID = col_integer(),
## TEACH_ID = col_integer(),
## gender = col_integer(),
## immigstatus = col_integer(),
## immigrant = col_integer(),
## school = col_integer(),
## race_r = col_integer(),
## pared = col_integer(),
## period_mean = col_integer(),
## group = col_integer(),
## focal = col_integer(),
## absent = col_integer(),
## ABSENT_7_9 = col_integer(),
## brainology = col_integer(),
## TUndermine = col_integer(),
## MS_Quarter1_SCI_7thgrade = col_character(),
## MS_Quarter2_SCI_7thgrade = col_character(),
## MS_Quarter3_SCI_7thgrade = col_character(),
## MS_Quarter4_SCI_7thgrade = col_character(),
## MS_Notes_grade7 = col_character()
## # ... with 108 more columns
## )
## See spec(...) for full column specifications.Creating variables
imuscle_esm$classroom<-paste(imuscle_esm$TEACHID, imuscle_esm$PERIOD)
imuscle_esm$uv_intervention<-ifelse(imuscle_esm$group == 2, 1,
ifelse(imuscle_esm$group == 4, 0, NA))
imuscle_student_survey$grade_binary <- ifelse(imuscle_student_survey$grade == 9, 1,
ifelse(imuscle_student_survey$grade == 7, 0, NA))
imuscle_student_survey$urm <- ifelse(imuscle_student_survey$race == 1, 0,
ifelse(imuscle_student_survey$race == 4, 0,
ifelse(imuscle_student_survey$race == 14, 0, 1)))
imuscle_student_survey$female <-ifelse(imuscle_student_survey$gender == 2, 1, 0)Join data
imuscle_all <- left_join(imuscle_student_survey, imuscle_esm, by = "STUDID")Predicting momentary interest
M1 <- lmer(INTEREST ~
female +
urm +
grade_binary +
uv_intervention +
t_week +
#t_week*uv_intervention +
t_week*SKILL*uv_intervention +
SKILL +
(1|STUDID) + (1|classroom), data = imuscle_all)
summary(M1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula:
## INTEREST ~ female + urm + grade_binary + uv_intervention + t_week +
## t_week * SKILL * uv_intervention + SKILL + (1 | STUDID) +
## (1 | classroom)
## Data: imuscle_all
##
## REML criterion at convergence: 7144
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.86922 -0.65608 -0.01205 0.64137 3.13681
##
## Random effects:
## Groups Name Variance Std.Dev.
## STUDID (Intercept) 0.25756 0.5075
## classroom (Intercept) 0.05714 0.2390
## Residual 0.65368 0.8085
## Number of obs: 2743, groups: STUDID, 292; classroom, 16
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 7.977e-01 1.502e-01 6.532e+01 5.313
## female -1.471e-01 6.821e-02 2.827e+02 -2.156
## urm 2.107e-01 8.230e-02 2.839e+02 2.561
## grade_binary -1.459e-01 1.423e-01 1.568e+01 -1.025
## uv_intervention 1.913e-01 1.775e-01 4.052e+01 1.078
## t_week -6.685e-03 7.117e-03 2.512e+03 -0.939
## SKILL 3.870e-01 3.601e-02 2.696e+03 10.747
## t_week:SKILL 5.297e-04 3.186e-03 2.523e+03 0.166
## uv_intervention:t_week -1.901e-02 1.000e-02 2.541e+03 -1.901
## uv_intervention:SKILL -5.465e-02 5.292e-02 2.711e+03 -1.033
## uv_intervention:t_week:SKILL 7.290e-03 4.684e-03 2.537e+03 1.556
## Pr(>|t|)
## (Intercept) 1.4e-06 ***
## female 0.0319 *
## urm 0.0110 *
## grade_binary 0.3209
## uv_intervention 0.2875
## t_week 0.3477
## SKILL < 2e-16 ***
## t_week:SKILL 0.8680
## uv_intervention:t_week 0.0574 .
## uv_intervention:SKILL 0.3018
## uv_intervention:t_week:SKILL 0.1197
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) female urm grd_bn uv_ntr t_week SKILL t_:SKI uv_n:_
## female -0.254
## urm -0.426 -0.013
## grade_binry -0.335 0.030 -0.049
## uv_intrvntn -0.522 0.007 0.037 -0.083
## t_week -0.371 0.004 0.010 0.018 0.303
## SKILL -0.471 0.027 0.002 0.009 0.389 0.573
## t_wek:SKILL 0.311 -0.006 -0.005 -0.002 -0.259 -0.888 -0.641
## uv_ntrvnt:_ 0.255 0.011 -0.003 -0.003 -0.447 -0.712 -0.408 0.632
## uv_nt:SKILL 0.306 -0.004 0.009 0.011 -0.576 -0.390 -0.680 0.436 0.603
## uv_:_:SKILL -0.205 -0.014 0.002 -0.001 0.377 0.604 0.436 -0.680 -0.884
## u_:SKI
## female
## urm
## grade_binry
## uv_intrvntn
## t_week
## SKILL
## t_wek:SKILL
## uv_ntrvnt:_
## uv_nt:SKILL
## uv_:_:SKILL -0.654ranova(M1)## ANOVA-like table for random-effects: Single term deletions
##
## Model:
## INTEREST ~ female + urm + grade_binary + uv_intervention + t_week +
## SKILL + (1 | STUDID) + (1 | classroom) + t_week:SKILL + uv_intervention:t_week +
## uv_intervention:SKILL + uv_intervention:t_week:SKILL
## npar logLik AIC LRT Df Pr(>Chisq)
## <none> 14 -3572.0 7172.0
## (1 | STUDID) 13 -3798.6 7623.2 453.18 1 < 2.2e-16 ***
## (1 | classroom) 13 -3597.5 7220.9 50.88 1 9.809e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1sjstats::icc(M1)##
## Linear mixed model
##
## Family : gaussian (identity)
## Formula: INTEREST ~ female + urm + grade_binary + uv_intervention + t_week + t_week * SKILL * uv_intervention + SKILL + (1 | STUDID) + (1 | classroom)
##
## ICC (STUDID): 0.2660
## ICC (classroom): 0.0590sjPlot::plot_model(M1, type = "int", swap.pred = FALSE)## Warning in (function (object, newdata = NULL, newparams = NULL, re.form =
## NULL, : unused arguments ignored
#konfound::konfound(M33_33, female)
#M33_33r<-r2glmm::r2beta(M33_33, method = "nsj")
#M33_33r
#MuMIn::r.squaredGLMM(M33_33)Predicting momentary boredom
M3 <- lmer(BORED ~
female +
urm +
grade_binary +
uv_intervention +
t_week +
#t_week*uv_intervention +
t_week*SKILL*uv_intervention +
SKILL +
(1|STUDID) + (1|classroom), data = imuscle_all)
summary(M3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: BORED ~ female + urm + grade_binary + uv_intervention + t_week +
## t_week * SKILL * uv_intervention + SKILL + (1 | STUDID) +
## (1 | classroom)
## Data: imuscle_all
##
## REML criterion at convergence: 7807.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.81565 -0.64217 -0.07159 0.68149 2.95309
##
## Random effects:
## Groups Name Variance Std.Dev.
## STUDID (Intercept) 0.40302 0.6348
## classroom (Intercept) 0.05301 0.2302
## Residual 0.83848 0.9157
## Number of obs: 2722, groups: STUDID, 292; classroom, 17
##
## Fixed effects:
## Estimate Std. Error df t value
## (Intercept) 1.729e+00 1.673e-01 8.396e+01 10.332
## female 1.176e-01 8.362e-02 2.819e+02 1.406
## urm -2.132e-01 1.008e-01 2.834e+02 -2.115
## grade_binary 1.151e-01 1.467e-01 1.575e+01 0.785
## uv_intervention -7.979e-02 1.909e-01 4.696e+01 -0.418
## t_week -4.221e-03 8.069e-03 2.428e+03 -0.523
## SKILL -1.767e-01 4.111e-02 2.658e+03 -4.297
## t_week:SKILL 1.390e-03 3.614e-03 2.491e+03 0.385
## uv_intervention:t_week 7.081e-03 1.134e-02 2.495e+03 0.624
## uv_intervention:SKILL -2.639e-02 6.049e-02 2.676e+03 -0.436
## uv_intervention:t_week:SKILL 3.867e-04 5.328e-03 2.507e+03 0.073
## Pr(>|t|)
## (Intercept) < 2e-16 ***
## female 0.1607
## urm 0.0353 *
## grade_binary 0.4444
## uv_intervention 0.6778
## t_week 0.6010
## SKILL 1.8e-05 ***
## t_week:SKILL 0.7006
## uv_intervention:t_week 0.5324
## uv_intervention:SKILL 0.6626
## uv_intervention:t_week:SKILL 0.9421
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) female urm grd_bn uv_ntr t_week SKILL t_:SKI uv_n:_
## female -0.276
## urm -0.467 -0.014
## grade_binry -0.315 0.035 -0.059
## uv_intrvntn -0.514 0.006 0.043 -0.081
## t_week -0.377 0.003 0.009 0.019 0.320
## SKILL -0.483 0.025 0.002 0.011 0.413 0.572
## t_wek:SKILL 0.317 -0.005 -0.004 -0.003 -0.274 -0.889 -0.638
## uv_ntrvnt:_ 0.258 0.012 -0.003 -0.001 -0.474 -0.711 -0.406 0.632
## uv_nt:SKILL 0.312 -0.001 0.009 0.013 -0.613 -0.388 -0.679 0.433 0.601
## uv_:_:SKILL -0.208 -0.015 0.001 -0.003 0.400 0.603 0.432 -0.678 -0.885
## u_:SKI
## female
## urm
## grade_binry
## uv_intrvntn
## t_week
## SKILL
## t_wek:SKILL
## uv_ntrvnt:_
## uv_nt:SKILL
## uv_:_:SKILL -0.650ranova(M3)## ANOVA-like table for random-effects: Single term deletions
##
## Model:
## BORED ~ female + urm + grade_binary + uv_intervention + t_week +
## SKILL + (1 | STUDID) + (1 | classroom) + t_week:SKILL + uv_intervention:t_week +
## uv_intervention:SKILL + uv_intervention:t_week:SKILL
## npar logLik AIC LRT Df Pr(>Chisq)
## <none> 14 -3903.6 7835.2
## (1 | STUDID) 13 -4180.7 8387.5 554.25 1 < 2.2e-16 ***
## (1 | classroom) 13 -3915.0 7856.1 22.82 1 1.783e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1sjstats::icc(M3)##
## Linear mixed model
##
## Family : gaussian (identity)
## Formula: BORED ~ female + urm + grade_binary + uv_intervention + t_week + t_week * SKILL * uv_intervention + SKILL + (1 | STUDID) + (1 | classroom)
##
## ICC (STUDID): 0.3113
## ICC (classroom): 0.0409sjPlot::plot_model(M3, type = "int", swap.pred = FALSE)## Warning in (function (object, newdata = NULL, newparams = NULL, re.form =
## NULL, : unused arguments ignored
#konfound::konfound(M33_33, female)
#M33_33r<-r2glmm::r2beta(M33_33, method = "nsj")
#M33_33r
#MuMIn::r.squaredGLMM(M33_33)Predicting post-interest
M2 <- lmer(interest3 ~
female +
urm +
grade_binary +
uv_intervention +
(1|classroom), data = imuscle_all)
summary(M2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: interest3 ~ female + urm + grade_binary + uv_intervention + (1 |
## classroom)
## Data: imuscle_all
##
## REML criterion at convergence: 9636.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.03393 -0.82010 0.09875 0.73611 2.73074
##
## Random effects:
## Groups Name Variance Std.Dev.
## classroom (Intercept) 0.3508 0.5923
## Residual 3.2335 1.7982
## Number of obs: 2390, groups: classroom, 17
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 4.57237 0.24407 16.66091 18.734 1.26e-12 ***
## female -0.38509 0.07474 2378.98346 -5.152 2.79e-07 ***
## urm -0.79683 0.08829 2380.79556 -9.025 < 2e-16 ***
## grade_binary -0.08791 0.31642 13.14210 -0.278 0.785
## uv_intervention 0.14941 0.30871 13.33579 0.484 0.636
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) female urm grd_bn
## female -0.160
## urm -0.278 -0.014
## grade_binry -0.435 0.012 -0.018
## uv_intrvntn -0.525 0.004 0.025 -0.091ranova(M2)## ANOVA-like table for random-effects: Single term deletions
##
## Model:
## interest3 ~ female + urm + grade_binary + uv_intervention + (1 |
## classroom)
## npar logLik AIC LRT Df Pr(>Chisq)
## <none> 7 -4818.2 9650.4
## (1 | classroom) 6 -4894.8 9801.6 153.24 1 < 2.2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1#sjstats::icc(M1)
#sjPlot::plot_model(M1, type = "int", swap.pred = FALSE)
#konfound::konfound(M33_33, female)
#M33_33r<-r2glmm::r2beta(M33_33, method = "nsj")
#M33_33r
#MuMIn::r.squaredGLMM(M33_33)