In: Statistics and Probability
The horsepower (Y, in bhp) of a motor car engine was measured at a chosen set of values of running speed (X, in rpm). The data are given below (the first row is the running speed in rpm and the second row is the horsepower in bhp):
rpm | 1100 | 1400 | 1700 | 2300 | 2700 | 3200 | 3500 | 4000 | 4300 | 4600 | 4900 | 5200 | 6100 |
Horsepower (bhp) | 41.11 | 50.41 | 89.98 | 97.76 | 143.29 | 168.77 | 171.62 | 199.89 | 207.23 | 233.34 | 241.96 | 252.73 | 311.44 |
The mean and sum of squares of the rpm are 3461.53853461.5385
rpm and 185040000.0000185040000.0000 rpm 22 respectively; the mean
of the horsepower values is 169.9638169.9638 bhp and the sum of the
products of the two variables is 9184823.00009184823.0000 rpm bhp.
A scatterplot displaying the data is shown below:
Please provide answers to the following to 3 decimals places where
appropriate:
Part a)
Compute the regression line for these data, and provide your
estimates of the slope and intercept parameters. To
calculate the slope and intercept please keep intermediate results
to 6 decimal places (when you calculate the intercept do not round
the value of the slope).
Slope:
Intercept:
Note: For the parts below, use the slope and intercept
values in Part a, corrected to 3 decimal places to
calculate answers.
Part b)
Based on the regression model, what level of horsepower would you
expect the engine to produce if running at 24002400 rpm?
Answer:
Part c)
Assuming the model you have fitted, if increase the running speed
by 100100 rpm, what would you expect the change in horsepower to
be?
Answer:
Part d)
The standard error of the estimate of the slope coefficient was
found to be 0.0016930.001693. Provide a 95% confidence interval for
the true underlying slope.
Confidence interval: ( , )
Part e)
Without extending beyond the existing range of speed values or
changing the number of observations, we would expect that
increasing the variance of the rpm speeds at which the horsepower
levels were found would make the confidence interval in (d)
A. narrower.
B. either wider or narrower depending on the
values chosen.
C. unchanged.
D. wider.
Part f)
If testing the null hypothesis that horsepower does not depend
linearly on rpm, what would be your test statistic? (For this part,
you are to calculate the test statistic by hand using appropriate
values from the answers you provided in part (a) accurate to 3
decimal places, and values given to you in part (d).)
Answer:
Part g)
Assuming the test is at the 1% significance level, what would you
conclude from the above hypothesis test?
A. Since the observed test statistic does not fall
in either the upper or lower 1 percentiles of the t distribution
with 1111 degrees of freedom, we cannot reject the null hypothesis
that the horsepower does not depend linearly on rpm.
B. Since the observed test statistic does not fall
in either the upper or lower 1/2 percentiles of the t distribution
with 1111 degrees of freedom, we can reject the null hypothesis
that the horsepower does not depend linearly on rpm.
C. Since the observed test statistic falls in
either the upper or lower 1/2 percentiles of the t distribution
with 1111 degrees of freedom, we cannot reject the null hypothesis
that the horsepower does not depend linearly on rpm.
D. Since the observed test statistic falls in
either the upper or lower 1/2 percentiles of the t distribution
with 1111 degrees of freedom, we can reject the null hypothesis
that the horsepower does not depend linearly on rpm.
E. Since the observed test statistic does not fall
in either the upper or lower 1/2 percentiles of the t distribution
with 1111 degrees of freedom, we cannot reject the null hypothesis
that the horsepower does not depend linearly on rpm.