Scribd
Upload
33 views9 pages

Code 2

The document analyzes delivery time data using ANOVA. It reads in a dataset with variables for crust, Coke, bread, and delivery time. It performs a one-way ANOVA comparing delivery times between treatments, finding a significant effect. A multi-factor ANOVA finds significant effects of crust, bread, and a crust-Coke interaction. Residuals are checked and pairwise comparisons are performed.

Uploaded by

AbhishekKumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
33 views9 pages

Code 2

The document analyzes delivery time data using ANOVA. It reads in a dataset with variables for crust, Coke, bread, and delivery time. It performs a one-way ANOVA comparing delivery times between treatments, finding a significant effect. A multi-factor ANOVA finds significant effects of crust, bread, and a crust-Coke interaction. Residuals are checked and pairwise comparisons are performed.

Uploaded by

AbhishekKumar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

ANOVA

Soumya Roy

Wed Jun 20 09:40:38 2018


# Reading the data file
d<-read.table("C:/Users/Soumya-
Roy/Desktop/DAR_18/Class_2/pizza.data",header=T)
d

## Crust Coke Bread Delivery


## 1 0 1 1 14
## 2 1 1 0 21
## 3 0 0 0 18
## 4 0 0 1 17
## 5 1 0 0 19
## 6 1 0 1 17
## 7 0 1 0 19
## 8 0 0 0 20
## 9 0 1 0 16
## 10 1 1 1 19
## 11 0 0 1 18
## 12 1 1 0 22
## 13 1 0 1 19
## 14 0 1 1 16
## 15 1 0 0 20
## 16 1 1 1 18

# Naming the variables


crust<-as.factor(d$Crust)
coke<-as.factor(d$Coke)
bread<-as.factor(d$Bread)
delivery<-d$Delivery

# Creating the treatments


treatment<-NULL
for(i in 1:16)
{
if(crust[i]=="0" & coke[i]=="0" & bread[i]=="0") treatment[i]<-"000"
if(crust[i]=="1" & coke[i]=="0" & bread[i]=="0") treatment[i]<-"100"
if(crust[i]=="0" & coke[i]=="1" & bread[i]=="0") treatment[i]<-"010"
if(crust[i]=="1" & coke[i]=="1" & bread[i]=="0") treatment[i]<-"110"
if(crust[i]=="0" & coke[i]=="0" & bread[i]=="1") treatment[i]<-"001"
if(crust[i]=="1" & coke[i]=="0" & bread[i]=="1") treatment[i]<-"101"
if(crust[i]=="0" & coke[i]=="1" & bread[i]=="1") treatment[i]<-"011"
if(crust[i]=="1" & coke[i]=="1" & bread[i]=="1") treatment[i]<-"111"
}
treatment<-as.factor(treatment)

# Strip Chart
stripchart(delivery~treatment,xlab="Treatment",ylab="Delivery
Time",vertical=T,col="red",pch=24,lwd=2)

# Plot of Average Delivery Times


library(gplots)

##
## Attaching package: 'gplots'

## The following object is masked from 'package:stats':


##
## lowess

plotmeans(delivery~treatment,xlab="Treatment",ylab="Delivery
Time",col="red",bars=T,barwidth=2,lwd=2)
# One-way ANOVA to check whether the average delivery times are significantly
different
mod_1<-aov(delivery~treatment)
summary(mod_1)

## Df Sum Sq Mean Sq F value Pr(>F)


## treatment 7 48.94 6.991 4.474 0.0259 *
## Residuals 8 12.50 1.563
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

# Multi-factor ANOVA
mod_2<-aov(delivery~crust*coke*bread)
summary(mod_2)

## Df Sum Sq Mean Sq F value Pr(>F)


## crust 1 18.062 18.062 11.56 0.00936 **
## coke 1 0.563 0.563 0.36 0.56511
## bread 1 18.062 18.062 11.56 0.00936 **
## crust:coke 1 10.562 10.562 6.76 0.03162 *
## crust:bread 1 0.062 0.062 0.04 0.84647
## coke:bread 1 1.563 1.563 1.00 0.34659
## crust:coke:bread 1 0.062 0.062 0.04 0.84647
## Residuals 8 12.500 1.563
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
# Multi-factor ANOVA after dropping the insignificant terms
mod_3<-aov(delivery~crust+coke+bread+crust:coke)
summary(mod_3)

## Df Sum Sq Mean Sq F value Pr(>F)


## crust 1 18.062 18.062 14.004 0.00325 **
## coke 1 0.563 0.563 0.436 0.52259
## bread 1 18.062 18.062 14.004 0.00325 **
## crust:coke 1 10.562 10.562 8.189 0.01547 *
## Residuals 11 14.188 1.290
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

# Plot of Residuals
plot(mod_3,which=c(1,2))
# Normality Check of Residuals
res<-mod_3$res
shapiro.test(res)
##
## Shapiro-Wilk normality test
##
## data: res
## W = 0.94095, p-value = 0.3608

# Homoscedasticity Check for Residuals


bartlett.test(res~crust)

##
## Bartlett test of homogeneity of variances
##
## data: res by crust
## Bartlett's K-squared = 0.82086, df = 1, p-value = 0.3649

bartlett.test(res~coke)

##
## Bartlett test of homogeneity of variances
##
## data: res by coke
## Bartlett's K-squared = 0.22757, df = 1, p-value = 0.6333

bartlett.test(res~bread)

##
## Bartlett test of homogeneity of variances
##
## data: res by bread
## Bartlett's K-squared = 0.20608, df = 1, p-value = 0.6499

# Plotting the Main Effects


library(gplots)
plotmeans(delivery~crust,xlab="Crust",ylab="Delivery
Time",col="red",bars=T,barwidth=2,lwd=2)
plotmeans(delivery~coke,xlab="Coke",ylab="Delivery
Time",col="red",bars=T,barwidth=2,lwd=2)
plotmeans(delivery~bread,xlab="Bread",ylab="Delivery
Time",col="red",bars=T,barwidth=2,lwd=2)

# Plotting the Interaction Effect


interaction.plot(crust,coke,delivery,main="crust*coke")
# Pairwise Comparison
TukeyHSD(mod_3)$bread

## diff lwr upr p adj


## 1-0 -2.125 -3.374809 -0.8751908 0.003253457

TukeyHSD(mod_3)$'crust:coke'

## diff lwr upr p adj


## 1:0-0:0 0.50 -1.9168119 2.91681188 0.92267036
## 0:1-0:0 -2.00 -4.4168119 0.41681188 0.11663937
## 1:1-0:0 1.75 -0.6668119 4.16681188 0.18879599
## 0:1-1:0 -2.50 -4.9168119 -0.08318812 0.04207525
## 1:1-1:0 1.25 -1.1668119 3.66681188 0.43969227
## 1:1-0:1 3.75 1.3331881 6.16681188 0.00324530

You might also like