The following data relate road width x and accident frequency y. Road width (in feet) was treated as the independent variable, and values y of the random variable Y, in accidents per 108 vehicle miles, were observed.

Assume that Y is normally distributed with mean A + Bx and constant variance for all x and that the sample is random. Interpolate if necessary.
(a) Fit a least-squares line to the data, and forecast the accident frequency when the road width is 55 feet.
(b) Construct a 95 percent prediction interval for Y+, a future observation of Y, corresponding to x+ = 55 feet.
(c) Suppose that two future observations on Y, both corresponding to x+ = 55 feet, are to be made. Construct prediction intervals for both of these observations so that the probability is at least 95 percent that both future values of Y will fall into them simultaneously. If k predictions are to be made, such as given in part (d), each with probability 1 €“ α, then the probability is at least 1 €“ kα that all k future observations will fall into their respective intervals.
(d) Construct a simultaneous tolerance interval for the future value of Y corresponding to x+ = 55 feet with P = 0.90 and 1 €“ α = 0.95.

Transcribed Image Text:

Nunber of Observations 7 26 30 92 85 78 81 y xi-354 50 x2 19,956 y 354516254 68 74 51 40 x 22,200