2021 H2 Mathematics Paper 2 Question 9

Linear Correlation and Regression

Answers

It is likely to provide a good model. The product moment correlation coefficient is close to 1,{1,} suggesting a strong linear correlation between t{t} and y.{y.} Moreover, the scatter diagram suggests that the rate of change of x{x} as t{t} increases is relatively constant, which is consistent with a linear model.
While relatively constant, there is still a slight increase in the rate of increase of x{x} so there may still be a better model nevertheless
(bi)
(bii)
c=2.29{c=2.29}
d=0.0573{d=0.0573}
r=0.999{r=0.999}
The slight increase in the rate of increase of x{x} as t{t} increases indicated by the scatter diagram suggests is more consistent with the second model.
This is also shown in the product moment correlation coefficient value of 0.999{0.999} which is closer to 1{1} (vs 0.9803{0.9803} for the first model), which indicates that there is a stronger linear correlation between lnx{\ln x} and t{t} as compared to the linear model.

Full solutions

(a)

It is likely to provide a good model. The product moment correlation coefficient is close to 1,{1,} suggesting a strong linear correlation between t{t} and y.{y.} Moreover, the scatter diagram suggests that the rate of change of x{x} as t{t} increases is relatively constant, which is consistent with a linear model.
While relatively constant, there is still a slight increase in the rate of increase of x{x} so there may still be a better model nevertheless{\blacksquare}
(bi)
(bii)
Using the GC,
c=2.29 (3 sf)  d=0.0573 (3 sf)  r=0.999 (3 sf)  \begin{align*} c &= 2.29 \textrm{ (3 sf)} \; \blacksquare \\ d &= 0.0573 \textrm{ (3 sf)} \; \blacksquare \\ r &= 0.999 \textrm{ (3 sf)} \; \blacksquare \end{align*}

(c)

The slight increase in the rate of increase of x{x} as t{t} increases indicated by the scatter diagram suggests is more consistent with the second model.
This is also shown in the product moment correlation coefficient value of 0.999{0.999} which is closer to 1{1} (vs 0.9803{0.9803} for the first model), which indicates that there is a stronger linear correlation between lnx{\ln x} and t{t} as compared to the linear model. {\blacksquare}