QUANT 632: Assignment 2
Jason Hoskin
2026-03-02
- Run libraries and import dataset.
## Run libraries
packages <- c("dplyr",
"psych",
"ggplot2",
"biotools",
"REdaS",
"mvShapiroTest",
"heplots",
"sjstats",
"agricolae")
installed <- packages %in% rownames(installed.packages())
if (any(!installed)) {
install.packages(packages[!installed])
}
lapply(packages, library, character.only = TRUE)## Warning in rgl.init(initValue, onlyNULL): X11 error: GLXBadContext## Warning: 'rgl.init' failed, will use the null device.
## See '?rgl.useNULL' for ways to avoid this warning.## Import dataset
url <- "https://github.com/JasonMHoskin/quant632-assignment2/raw/refs/heads/main/data_assn2.rda"
download.file(url, destfile = "file.RDA", mode = "wb")
load("file.RDA")
df1 <- assn2
## Remove unnecessary variables
rm(installed, packages, url)- Recode IV (WORKSTAT): 0 = homemakers, 1 = working women
## Recode
df1$workstat <- ifelse(df1$workstat == 1, 1, 0)
## Verify recoding is accurate
df1 |>
group_by(workstat) |>
count(workstat)## # A tibble: 2 × 2
## # Groups: workstat [2]
## workstat n
## <dbl> <int>
## 1 0 137
## 2 1 244assn2 |>
group_by(workstat) |>
count(workstat)## # A tibble: 2 × 2
## # Groups: workstat [2]
## workstat n
## <hvn_lbll> <int>
## 1 1 244
## 2 2 137- Generate “grouping” variable (The number 1 should be assigned to working women without children, 2 to working women with children, 3 to homemakers without children, and 4 to homemakers with children) and sort cases by the “grouping” variable.
## Make IVs factors
df1$workstat <- as.factor(df1$workstat)
df1$children <- as.factor(df1$children)
## Generate grouping variable
df1 <- df1 |>
mutate(
grouping = case_when(
df1$workstat == 0 & df1$children == 0 ~ 1L,
df1$workstat == 1 & df1$children == 0 ~ 2L,
df1$workstat == 0 & df1$children == 1 ~ 3L,
df1$workstat == 1 & df1$children == 1 ~ 4L,
TRUE ~ NA_integer_
)
)
df1$grouping <- as.factor(df1$grouping)
df1 <- arrange(df1, grouping)
head(df1)## subno workstat children control attmar attrole atthouse grouping
## 1 4 0 0 6 24 31 28 1
## 2 7 0 0 6 27 44 30 1
## 3 8 0 0 8 18 48 24 1
## 4 12 0 0 6 16 45 22 1
## 5 14 0 0 6 20 45 25 1
## 6 23 0 0 6 21 30 28 1- Run frequencies for DVs based on the grouping variable.
describeBy(df1, group = df1$grouping)##
## Descriptive statistics by group
## group: 1
## vars n mean sd median trimmed mad min max range skew
## subno 1 49 327.53 245.35 280 318.95 324.69 4 757 753 0.27
## workstat 2 49 1.00 0.00 1 1.00 0.00 1 1 0 NaN
## children 3 49 1.00 0.00 1 1.00 0.00 1 1 0 NaN
## control 4 49 6.80 1.41 7 6.71 1.48 5 10 5 0.57
## attmar 5 49 20.61 5.86 20 20.12 5.93 12 35 23 0.76
## attrole 6 49 37.16 6.66 37 37.29 7.41 24 49 25 -0.11
## atthouse 7 49 22.80 4.09 23 22.80 4.45 14 31 17 0.00
## grouping 8 49 1.00 0.00 1 1.00 0.00 1 1 0 NaN
## kurtosis se
## subno -1.34 35.05
## workstat NaN 0.00
## children NaN 0.00
## control -0.64 0.20
## attmar 0.05 0.84
## attrole -1.00 0.95
## atthouse -0.75 0.58
## grouping NaN 0.00
## ------------------------------------------------------------
## group: 2
## vars n mean sd median trimmed mad min max range skew
## subno 1 55 339.51 179.14 359 336.36 220.91 65 690 625 0.01
## workstat 2 55 2.00 0.00 2 2.00 0.00 2 2 0 NaN
## children 3 55 1.00 0.00 1 1.00 0.00 1 1 0 NaN
## control 4 55 6.75 1.44 7 6.62 1.48 5 10 5 0.62
## attmar 5 53 24.57 8.22 25 24.51 7.41 11 44 33 0.15
## attrole 6 55 31.24 6.34 31 31.24 5.93 18 43 25 0.04
## atthouse 7 54 23.54 4.41 24 23.57 5.19 14 33 19 -0.09
## grouping 8 55 2.00 0.00 2 2.00 0.00 2 2 0 NaN
## kurtosis se
## subno -1.25 24.16
## workstat NaN 0.00
## children NaN 0.00
## control -0.50 0.19
## attmar -0.94 1.13
## attrole -0.87 0.85
## atthouse -0.95 0.60
## grouping NaN 0.00
## ------------------------------------------------------------
## group: 3
## vars n mean sd median trimmed mad min max range skew
## subno 1 88 328.36 192.47 319.5 328.35 255.01 5 758 753 0.04
## workstat 2 88 1.00 0.00 1.0 1.00 0.00 1 1 0 NaN
## children 3 88 2.00 0.00 2.0 2.00 0.00 2 2 0 NaN
## control 4 87 6.54 1.25 6.0 6.44 1.48 5 10 5 0.55
## attmar 5 88 20.57 7.01 19.0 19.78 5.93 11 44 33 1.12
## attrole 6 88 37.26 6.32 38.0 37.21 7.41 26 55 29 0.19
## atthouse 7 88 22.33 3.75 22.0 22.33 2.97 11 32 21 -0.12
## grouping 8 88 3.00 0.00 3.0 3.00 0.00 3 3 0 NaN
## kurtosis se
## subno -0.95 20.52
## workstat NaN 0.00
## children NaN 0.00
## control -0.46 0.13
## attmar 1.12 0.75
## attrole -0.37 0.67
## atthouse 0.39 0.40
## grouping NaN 0.00
## ------------------------------------------------------------
## group: 4
## vars n mean sd median trimmed mad min max range skew
## subno 1 189 295.25 196.47 294 288.06 268.35 2 754 752 0.20
## workstat 2 189 2.00 0.00 2 2.00 0.00 2 2 0 NaN
## children 3 189 2.00 0.00 2 2.00 0.00 2 2 0 NaN
## control 4 189 6.71 1.17 7 6.66 1.48 5 10 5 0.35
## attmar 5 187 23.18 8.72 21 22.25 7.41 11 53 42 0.94
## attrole 6 189 34.49 6.96 35 34.35 7.41 18 52 34 0.14
## atthouse 7 189 23.90 4.44 24 23.91 4.45 12 34 22 -0.03
## grouping 8 189 4.00 0.00 4 4.00 0.00 4 4 0 NaN
## kurtosis se
## subno -1.12 14.29
## workstat NaN 0.00
## children NaN 0.00
## control -0.61 0.09
## attmar 0.36 0.64
## attrole -0.47 0.51
## atthouse -0.39 0.32
## grouping NaN 0.00- Check for missing data. If there are, report the number of missing data by variable. Delete cases with missing data.
## Check for NAs
colSums(is.na(df1))## subno workstat children control attmar attrole atthouse grouping
## 0 0 0 1 4 0 1 0## Remove NAs
df2 <- na.omit(df1)
## Confirm NAs are removed
colSums(is.na(df2))## subno workstat children control attmar attrole atthouse grouping
## 0 0 0 0 0 0 0 0- Draw a histogram of each DV by the grouping variable and visually evaluate if there are potential outliers, if the distributions are normal, and if the groups could be homogenous in variances.
## Create histograms
ggplot(df2, aes(x = control)) +
geom_histogram(
aes(y = after_stat(density)),
binwidth = 1,
colour = "black",
fill = "grey"
) +
labs(title = "locus of control by group") +
facet_wrap( ~ grouping) +
theme_grey(base_size = 25) +
geom_density(alpha = .3, fill = "skyblue")
ggplot(df2, aes(x = attmar)) +
geom_histogram(
aes(y = after_stat(density)),
binwidth = 1,
colour = "black",
fill = "grey"
) +
labs(title = "attitudes towrad current marital status by group") +
facet_wrap( ~ grouping) +
theme_grey(base_size = 25) +
geom_density(alpha = .3, fill = "skyblue")
ggplot(df2, aes(x = attrole)) +
geom_histogram(
aes(y = after_stat(density)),
binwidth = 1,
colour = "black",
fill = "grey"
) +
labs(title = "attitudes toward women's role by group") +
facet_wrap( ~ grouping) +
theme_grey(base_size = 25) +
geom_density(alpha = .3, fill = "skyblue")
ggplot(df2, aes(x = atthouse)) +
geom_histogram(
aes(y = after_stat(density)),
binwidth = 1,
colour = "black",
fill = "grey"
) +
labs(title = "attitudes toward housework by group") +
facet_wrap( ~ grouping) +
theme_grey(base_size = 25) +
geom_density(alpha = .3, fill = "skyblue")
- Check for univariate outliers (use criterion: z = ±3.3). If there are, report the subject numbers (SUBNO) of the univariate outliers. Delete the cases with univariate outliers.
## Check for univariate outliers
df2$z_control <- scale(df2$control)
df2 |>
filter(z_control >= 3.3 | z_control <= -3.3) |>
pull(subno)## Warning: Using one column matrices in `filter()` was deprecated in dplyr 1.1.0.
## ℹ Please use one dimensional logical vectors instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## numeric(0)df2$z_attmar <- scale(df2$attmar)
df2 |>
filter(z_attmar >= 3.3 | z_attmar <= -3.3) |>
pull(subno)## [1] 10 570df2$z_attrole <- scale(df2$attrole)
df2 |>
filter(z_attrole >= 3.3 | z_attmar <= -3.3) |>
pull(subno)## numeric(0)df2$z_atthouse <- scale(df2$atthouse)
df2 |>
filter(z_atthouse >= 3.3 | z_atthouse <= -3.3) |>
pull(subno)## numeric(0)## Remove univariate outliers
df3 <- df2 |>
filter(
subno != 10 & subno != 150
)- Check for multivariate outliers (criterion: χ2 at α = .001 using Mahalanobis distance). Show the Q-Q plot. If there are, report the subject numbers (SUBNO) of the multivariate outliers. Delete the cases with multivariate outliers if there are.
## Subset dataset
worknochild <- subset(df3, grouping == 1)
workchild <- subset(df3, grouping == 2)
homenochild <- subset(df3, grouping == 3)
homechild <- subset(df3, grouping == 4)
## Calculate cutoff distance
alpha <- .001
cutoff <- (qchisq(p = 1 - alpha, df = 4))
## Calculate multivariate variables
worknochild$mahal0 <- outlier(worknochild[, c(4:7)], plot = T)
worknochild$outlier0 <- ifelse(worknochild$mahal0 > cutoff, 1, 0)
table(worknochild$outlier0)##
## 0
## 49workchild$mahal0 <- outlier(workchild[, c(4:7)], plot = T)
workchild$outlier0 <- ifelse(workchild$mahal0 > cutoff, 1, 0)
table(workchild$outlier0)##
## 0
## 52homenochild$mahal0 <- outlier(homenochild[, c(4:7)], plot = T)
homenochild$outlier0 <- ifelse(homenochild$mahal0 > cutoff, 1, 0)
table(homenochild$outlier0)##
## 0
## 86homechild$mahal0 <- outlier(homechild[, c(4:7)], plot = T)
homechild$outlier0 <- ifelse(homechild$mahal0 > cutoff, 1, 0)
table(homechild$outlier0)##
## 0
## 186## Remove multivariate outliers and remove unnecessary variables
df4 <- df3 |>
filter(!row_number() %in% c(0, 49, 52, 86, 186)) |>
dplyr::select(subno,
grouping,
workstat,
children,
control,
attmar,
attrole,
atthouse)
rm(homechild, homenochild, workchild, worknochild, alpha, cutoff)- Evaluate whether or not the dataset meets the assumption of homogeneity of variance-covariance matrices at .001 level.
boxM(df4[, 5:8], df4[, 2])##
## Box's M-test for Homogeneity of Covariance Matrices
##
## data: df4[, 5:8] by df4[, 2]
## Chi-Sq (approx.) = 47.0621, df = 30, p-value = 0.02454- Evaluate whether or not the dataset meets the assumption of substantial correlations among DVs at .05 level.
bart_spher(df4[, 5:8])## Bartlett's Test of Sphericity
##
## Call: bart_spher(x = df4[, 5:8])
##
## X2 = 97.74
## df = 6
## p-value < 2.22e-16- Evaluate the possibility of multicollinearity.
cor(df4[, 5:8])## control attmar attrole atthouse
## control 1.00000000 0.17581370 0.02726253 0.1319919
## attmar 0.17581370 1.00000000 -0.09973102 0.3181755
## attrole 0.02726253 -0.09973102 1.00000000 -0.3317072
## atthouse 0.13199195 0.31817546 -0.33170724 1.0000000cor.plot(cor(df4[, 5:8]))
- Interpret the results of MANOVA (Use α = .01 for each single test. Remember we have 4 DVs. We want to use an adjusted level of significance for each hypothesis testing, i.e., .01.). Were the combined DVs significantly affected by women’s work status, having/not having children, and their interaction?
fit1.1 <- manova(cbind(control, attmar, attrole, atthouse) ~ workstat * children, data = df4)
summary(fit1.1, test = "Wilks")## Df Wilks approx F num Df den Df Pr(>F)
## workstat 1 0.92703 7.1237 4 362 1.572e-05 ***
## children 1 0.98376 1.4937 4 362 0.2035
## workstat:children 1 0.98080 1.7716 4 362 0.1339
## Residuals 365
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(fit1.1, test = "Pillai")## Df Pillai approx F num Df den Df Pr(>F)
## workstat 1 0.072971 7.1237 4 362 1.572e-05 ***
## children 1 0.016237 1.4937 4 362 0.2035
## workstat:children 1 0.019200 1.7716 4 362 0.1339
## Residuals 365
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(fit1.1, test = "Hotelling")## Df Hotelling-Lawley approx F num Df den Df Pr(>F)
## workstat 1 0.078714 7.1237 4 362 1.572e-05 ***
## children 1 0.016505 1.4937 4 362 0.2035
## workstat:children 1 0.019576 1.7716 4 362 0.1339
## Residuals 365
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary(fit1.1, test = "Roy")## Df Roy approx F num Df den Df Pr(>F)
## workstat 1 0.078714 7.1237 4 362 1.572e-05 ***
## children 1 0.016505 1.4937 4 362 0.2035
## workstat:children 1 0.019576 1.7716 4 362 0.1339
## Residuals 365
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1summary.aov(fit1.1)## Response control :
## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 0.24 0.24339 0.1508 0.6980
## children 1 0.55 0.55206 0.3421 0.5590
## workstat:children 1 1.38 1.37997 0.8552 0.3557
## Residuals 365 588.98 1.61366
##
## Response attmar :
## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 641.6 641.59 10.5984 0.001238 **
## children 1 49.1 49.13 0.8117 0.368224
## workstat:children 1 31.0 30.97 0.5117 0.474874
## Residuals 365 22095.6 60.54
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Response attrole :
## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 879.4 879.39 19.3899 1.402e-05 ***
## children 1 192.6 192.60 4.2467 0.04003 *
## workstat:children 1 160.1 160.12 3.5305 0.06104 .
## Residuals 365 16553.9 45.35
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Response atthouse :
## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 137.1 137.054 7.4952 0.006489 **
## children 1 0.1 0.103 0.0057 0.940121
## workstat:children 1 8.2 8.237 0.4505 0.502535
## Residuals 365 6674.3 18.286
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1## Test for multivariate normality
worknochild <- subset(df4, grouping == 1)
workchild <- subset(df4, grouping == 2)
homenochild <- subset(df4, grouping == 3)
homechild <- subset(df4, grouping == 4)
mat <- as.matrix(df4[, 5:8])
mat1 <- as.matrix(worknochild[, 5:8])
mat2 <- as.matrix(workchild[, 5:8])
mat3 <- as.matrix(homenochild[, 5:8])
mat4 <- as.matrix(homechild[, 5:8])
# Run Shapiro test
mvShapiro.Test(mat)##
## Generalized Shapiro-Wilk test for Multivariate Normality by
## Villasenor-Alva and Gonzalez-Estrada
##
## data: mat
## MVW = 0.97343, p-value < 2.2e-16mvShapiro.Test(mat1)##
## Generalized Shapiro-Wilk test for Multivariate Normality by
## Villasenor-Alva and Gonzalez-Estrada
##
## data: mat1
## MVW = 0.96648, p-value = 0.08084mvShapiro.Test(mat2)##
## Generalized Shapiro-Wilk test for Multivariate Normality by
## Villasenor-Alva and Gonzalez-Estrada
##
## data: mat2
## MVW = 0.95727, p-value = 0.005707mvShapiro.Test(mat3)##
## Generalized Shapiro-Wilk test for Multivariate Normality by
## Villasenor-Alva and Gonzalez-Estrada
##
## data: mat3
## MVW = 0.96196, p-value = 3.874e-05mvShapiro.Test(mat4)##
## Generalized Shapiro-Wilk test for Multivariate Normality by
## Villasenor-Alva and Gonzalez-Estrada
##
## data: mat4
## MVW = 0.96667, p-value = 3.744e-11cqplot(fit1.1)
res <- residuals(fit1.1)
d2 <- mahalanobis(res, colMeans(res), cov(res))
qqplot(
qchisq(ppoints(length(d2)), df = 4),
sort(d2),
xlab = expression(chi^2 ~ quantiles),
ylab = expression("Residual Mahalanobis Distance" * D^2)
)
abline(0, 1)
12.Conduct the post-hoc univariate test with Bonferroni adjustment.
alpha <- .01
## Post-hoc locus of control
control1 <- aov(control ~ grouping, data=df4)
hsd1 <- HSD.test(control1, "grouping", console = TRUE, alpha = alpha)##
## Study: control1 ~ "grouping"
##
## HSD Test for control
##
## Mean Square Error: 1.613656
##
## grouping, means
##
## control std r se Min Max Q25 Q50 Q75
## 1 6.812500 1.424053 48 0.18335167 5 10 6 7 8
## 2 6.680000 1.463013 50 0.17964721 5 10 6 6 7
## 3 6.564706 1.248304 85 0.13778312 5 10 6 6 7
## 4 6.715054 1.180696 186 0.09314274 5 10 6 7 8
##
## Alpha: 0.01 ; DF Error: 365
## Critical Value of Studentized Range: 4.433438
##
## Groups according to probability of means differences and alpha level( 0.01 )
##
## Treatments with the same letter are not significantly different.
##
## control groups
## 1 6.812500 a
## 4 6.715054 a
## 2 6.680000 a
## 3 6.564706 a## Post-hoc attmar
attmar1 <- aov(attmar ~ grouping, data=df4)
hsd2 <- HSD.test(attmar1, "grouping", console = TRUE, alpha = alpha)##
## Study: attmar1 ~ "grouping"
##
## HSD Test for attmar
##
## Mean Square Error: 60.53597
##
## grouping, means
##
## attmar std r se Min Max Q25 Q50 Q75
## 1 20.62500 5.923717 48 1.1230165 12 35 16.0 20 24.25
## 2 24.44000 7.759550 50 1.1003269 12 39 18.5 25 30.00
## 3 20.54118 7.107261 85 0.8439122 11 44 16.0 19 24.00
## 4 23.01613 8.463898 186 0.5704929 11 49 16.0 21 28.00
##
## Alpha: 0.01 ; DF Error: 365
## Critical Value of Studentized Range: 4.433438
##
## Groups according to probability of means differences and alpha level( 0.01 )
##
## Treatments with the same letter are not significantly different.
##
## attmar groups
## 2 24.44000 a
## 4 23.01613 a
## 1 20.62500 a
## 3 20.54118 a## Post-house attrole
attrole1 <- aov(attrole ~ grouping, data=df4)
hsd3 <- HSD.test(attrole1, "grouping", console = TRUE, alpha = alpha)##
## Study: attrole1 ~ "grouping"
##
## HSD Test for attrole
##
## Mean Square Error: 45.35308
##
## grouping, means
##
## attrole std r se Min Max Q25 Q50 Q75
## 1 37.20833 6.725494 48 0.9720370 24 49 32 37.5 43.25
## 2 31.60000 6.430159 50 0.9523978 18 43 27 32.0 36.75
## 3 37.15294 6.404065 85 0.7304558 26 55 32 37.0 42.00
## 4 34.59140 6.958155 186 0.4937953 18 52 29 35.0 39.00
##
## Alpha: 0.01 ; DF Error: 365
## Critical Value of Studentized Range: 4.433438
##
## Groups according to probability of means differences and alpha level( 0.01 )
##
## Treatments with the same letter are not significantly different.
##
## attrole groups
## 1 37.20833 a
## 3 37.15294 a
## 4 34.59140 ab
## 2 31.60000 b## Post-hoc atthouse
atthouse1 <- aov(atthouse ~ grouping, data=df4)
hsd4 <- HSD.test(atthouse1, "grouping", console = TRUE, alpha = alpha)##
## Study: atthouse1 ~ "grouping"
##
## HSD Test for atthouse
##
## Mean Square Error: 18.28567
##
## grouping, means
##
## atthouse std r se Min Max Q25 Q50 Q75
## 1 22.79167 4.135772 48 0.6172126 14 31 19.75 23 26
## 2 23.58000 4.562894 50 0.6047424 14 33 19.25 24 27
## 3 22.36471 3.766541 85 0.4638163 11 32 20.00 22 25
## 4 23.84409 4.446972 186 0.3135443 12 34 21.00 24 27
##
## Alpha: 0.01 ; DF Error: 365
## Critical Value of Studentized Range: 4.433438
##
## Groups according to probability of means differences and alpha level( 0.01 )
##
## Treatments with the same letter are not significantly different.
##
## atthouse groups
## 4 23.84409 a
## 2 23.58000 a
## 1 22.79167 a
## 3 22.36471 a- What is the effect size of each of the significant effects? Compute partial eta squared up to three decimal points. You can use manual computations (show your calculations) or use the sjstats package.
fit2.1 <- aov(control ~ workstat, data=df4)
summary(fit2.1)## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 0.2 0.2434 0.151 0.698
## Residuals 367 590.9 1.6101effectsize::eta_squared(fit2.1, partial = TRUE)## For one-way between subjects designs, partial eta squared is equivalent
## to eta squared. Returning eta squared.## # Effect Size for ANOVA
##
## Parameter | Eta2 | 95% CI
## -----------------------------------
## workstat | 4.12e-04 | [0.00, 1.00]
##
## - One-sided CIs: upper bound fixed at [1.00].fit2.2 <- aov(attmar ~ workstat, data=df4)
summary(fit2.2)## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 642 641.6 10.62 0.00122 **
## Residuals 367 22176 60.4
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1effectsize::eta_squared(fit2.2, partial = TRUE)## For one-way between subjects designs, partial eta squared is equivalent
## to eta squared. Returning eta squared.## # Effect Size for ANOVA
##
## Parameter | Eta2 | 95% CI
## -------------------------------
## workstat | 0.03 | [0.01, 1.00]
##
## - One-sided CIs: upper bound fixed at [1.00].fit2.3 <- aov(attrole ~ workstat, data=df4)
summary(fit2.3)## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 879 879.4 19.09 1.63e-05 ***
## Residuals 367 16907 46.1
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1effectsize::eta_squared(fit2.3, partial = TRUE)## For one-way between subjects designs, partial eta squared is equivalent
## to eta squared. Returning eta squared.## # Effect Size for ANOVA
##
## Parameter | Eta2 | 95% CI
## -------------------------------
## workstat | 0.05 | [0.02, 1.00]
##
## - One-sided CIs: upper bound fixed at [1.00].fit2.4 <- aov(atthouse ~ workstat, data=df4)
summary(fit2.4)## Df Sum Sq Mean Sq F value Pr(>F)
## workstat 1 137 137.05 7.527 0.00638 **
## Residuals 367 6683 18.21
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1effectsize::eta_squared(fit2.4, partial = TRUE)## For one-way between subjects designs, partial eta squared is equivalent
## to eta squared. Returning eta squared.## # Effect Size for ANOVA
##
## Parameter | Eta2 | 95% CI
## -------------------------------
## workstat | 0.02 | [0.00, 1.00]
##
## - One-sided CIs: upper bound fixed at [1.00].