Data 606 Homework 8

(a) Predicted baby weight = 120.07 - 1.93 * parity

(b) Parody 1 = 120.07 - 1.93 or 118.14 / Parody 0 = 120.07 / There is a 1.93 birthweight difference between first borns and latters.

(c) There is not a statistically signifcant relationship, I infer this from the p-value being greater than .05.

(a) Predicted absenteeism = 18.93 - 9.11 * eth + 3.10 * sex + 2.15 * lrn

(b) Slow learners are absent 2.15 more days than the average leaner. non-Aboriginals are absent 9.11 days more than Aboriginals. Males are absent 3.1 days more than females.

(c) residual = 2 - 24.18 = -22.18

(d)

observations = 146
variablesC = 3
resVariance = 240.57
absenteeVar = 264.17
rs = 1-(resVariance/absenteeVar)
ars = 1 -(resVariance/absenteeVar) * ( (observations-1) / (observations-variablesC-1) )
rs

## [1] 0.08933641

ars

## [1] 0.07009704

(a) The learner variable should be removed first because the residuals show the largest variance.

(a) We can see that if a mission took place below 64 degrees, it has at least one damaged O-ring. There seems to be a correlation between temperature and damage to O-rings. That as temperatures go higher, O-ring damage decreases.

(b) Temperature being the explanatory variable displays a negative relationship with O-ring damage. The p-value is so close to 0, that we can induce this is not due to chance.

(c)

(d) Well, only using the model “yes,” due to the p-value. However I imagine as the missions incremented, technology regarding their safety and durability had increased as well. With a little more information, “time” could have been our explanatory variable.

###(a)

logModel = function(temperature){
  oRing  = 11.6630 - 0.2162 * temperature
  answer = 100*(exp(oRing) / (1+exp(oRing)))
  return(answer)
}

temps = c(57,65,59,67,61,69,63,71,51,53,55)
probabilites = sapply(temps,logModel)
probabilites

##  [1] 34.064976  8.393843 25.109139  5.612566 17.869707  3.715479 12.372702
##  [8]  2.443024 65.402974 55.092283 44.324565

(b)

plot(y=probabilites, x=temps)
curve(logModel(x), from=40,to=80,add=TRUE,xlab='o-ring damage',ylab='temperature')

(c) As explained before, as temperature rises, the technology used on incremental missions increases, but we do not account for that. It may be that temperature has NO affect on o-rings; but we wouldn’t know with this little information.

Data 606 Homework 8

Michael Muller

May 7, 2017

(a) Predicted baby weight = 120.07 - 1.93 * parity

(b) Parody 1 = 120.07 - 1.93 or 118.14 / Parody 0 = 120.07 / There is a 1.93 birthweight difference between first borns and latters.

(c) There is not a statistically signifcant relationship, I infer this from the p-value being greater than .05.

(a) Predicted absenteeism = 18.93 - 9.11 * eth + 3.10 * sex + 2.15 * lrn

(b) Slow learners are absent 2.15 more days than the average leaner. non-Aboriginals are absent 9.11 days more than Aboriginals. Males are absent 3.1 days more than females.

(c) residual = 2 - 24.18 = -22.18

(d)

(a) The learner variable should be removed first because the residuals show the largest variance.

(a) We can see that if a mission took place below 64 degrees, it has at least one damaged O-ring. There seems to be a correlation between temperature and damage to O-rings. That as temperatures go higher, O-ring damage decreases.

(b) Temperature being the explanatory variable displays a negative relationship with O-ring damage. The p-value is so close to 0, that we can induce this is not due to chance.

(c)

(d) Well, only using the model “yes,” due to the p-value. However I imagine as the missions incremented, technology regarding their safety and durability had increased as well. With a little more information, “time” could have been our explanatory variable.

(b)

(c) As explained before, as temperature rises, the technology used on incremental missions increases, but we do not account for that. It may be that temperature has NO affect on o-rings; but we wouldn’t know with this little information.