R Assignment 2

load(“parenthood.Rdata”)

scatterplot( parenthood$dan.sleep,
      parenthood$dan.grump,
      main = "Dan's grumpiness per hours slept",
      xlab = "Dan sleep (hours)",
      ylab = "Dan grumpiness (0-100)",
      xlim = c(0,14),   # scale the x-axis
      ylim = c(0,100),    # scale the y-axis
      pch = 20,         # change the plot symbol
      frame.plot = FALSE, # don’t draw a box
      col= "red"
)

The correlation coefficient is what I expected because it shows a negative slope with a strong linear relationship.

cor(parenthood$dan.sleep,parenthood$dan.grump)

[1] -0.903384

lm.grump.sleep <- lm(parenthood$dan.grump ~ parenthood$dan.sleep)
# We look at the residuals:
plot(lm.grump.sleep, which = 1)  

# We look at the QQplot
plot(lm.grump.sleep, which = 2)

Yes because the plots all fit one the same line.

scatterplot( parenthood$baby.sleep,
      parenthood$dan.grump,
      main = "Dan's grunpiness per hours Baby slept",
      xlab = "Baby Sleep (hours)",
      ylab = "Dan grumpiness (0-100)",
      xlim = c(0,14),   # scale the x-axis
      ylim = c(0,100),    # scale the y-axis
      pch = 20,         # change the plot symbol
      frame.plot = FALSE, # don’t draw a box
      col= "blue"
)

The correlation coefficient is what I expected because it shows a negative slope with a fairly strong linear relationship.

cor(parenthood$baby.sleep,parenthood$dan.grump)

[1] -0.5659637

lm.2nd <- lm(parenthood$dan.grump ~ parenthood$baby.sleep)
# We look at the residuals:
plot(lm.2nd, which = 1)  

# We look at the QQplot
plot(lm.2nd, which = 2)

Yes because the plots all fit one the same line.

scatterplot( parenthood$baby.sleep,
      parenthood$dan.sleep,
      main = "Amount Dan sleep when Baby sleep",
      xlab = "Baby hours slept (hours)",
      ylab = "Dan hours slept (0-100)",
      xlim = c(0,14),   # scale the x-axis
      ylim = c(0,14),    # scale the y-axis
      pch = 20,         # change the plot symbol
      frame.plot = FALSE, # don’t draw a box
      col= "green"
)

The correlation coefficient is what I expected because of the postiive slope with the fairly strong linear relationship.

cor(parenthood$baby.sleep,parenthood$dan.sleep)

lm.3rd <- lm(parenthood$dan.sleep ~ parenthood$baby.sleep)
# We look at the residuals:
plot(lm.3rd, which = 1) 

# We look at the QQplot
plot(lm.3rd, which = 2)

Yes because the plots all fit one the same line.

The scatterplot with the strongest correlation would be Dan.sleep predicting Dan.grump. The more sleep Dan has, the less grumpy he is, shown by the negative slope with the strong linear relationship with its r being closest to 1.