In: Statistics and Probability
Keeping water supplies clean requires regular measurement of levels of pollutants. The measurements are indirecta typical analysis involves forming a dye by a chemical reaction with the dissolved pollutant, then passing light through the solution and measuring its "absorbence." To calibrate such measurements, the laboratory measures known standard solutions and uses regression to relate absorbence and pollutant concentration. This is usually done every day. Here is one series of data on the absorbence for different levels of nitrates. Nitrates are measured in milligrams per liter of water.
Nitrates | 50 | 50 | 100 | 200 | 400 | 800 | 1200 | 1600 | 2000 | 2000 |
Absorbence | 7.0 | 7.4 | 12.9 | 24.0 | 47.0 | 94.0 | 138.0 | 183.0 | 229.0 | 226.0 |
(a) Chemical theory says that these data should lie on a
straight line and if the correlation is not at least 0.997 then the
calibration procedure is repeated. Find the correlation. (Use 4
decimal places.)
r =
(b) Must the calibration be done again?
YesNo
(c) The calibration process sets nitrate level and measures
absorbence. Once established, the linear relationship will be used
to estimate the nitrate level in water from a measurement of
absorbance. What is the equation of the line used for estimation?
(Use 2 decimal places for intercept and 3 decimal places for
slope.)
= + x
(d) What is the estimated nitrate level in a water specimen with
absorbence 39? (Use 1 decimal place.)
mg/l
(e) Do you expect estimates of nitrate level from absorbence to be
quite accurate?
This prediction should be very inaccurate because the
relationship is too perfectly linear.
This prediction is of a value that is not in the range of the data
and therefore cannot be accurate.
This prediction should be very accurate because the relationship is
so strong as indicated by r.
Since the calibration is so important, it is inaccurate to use this
regression to predict.
Nitrates (y) | Absorbence(x) | xy | x2 | y2 | |
50 | 7 | 350 | 49 | 2500 | |
50 | 7.4 | 370 | 54.76 | 2500 | |
100 | 12.9 | 1290 | 166.41 | 10000 | |
200 | 24 | 4800 | 576 | 40000 | |
400 | 47 | 18800 | 2209 | 160000 | |
800 | 94 | 75200 | 8836 | 640000 | |
1200 | 138 | 165600 | 19044 | 1440000 | |
1600 | 183 | 292800 | 33489 | 2560000 | |
2000 | 229 | 458000 | 52441 | 4000000 | |
2000 | 226 | 452000 | 51076 | 4000000 | |
Sum | 8400 | 968.3 | 1469210 | 167941.17 | 12855000 |
a)
Formula for correlation coefficient is
where sx and sy are the sample standard deviations, and sxy is the sample covariance.
= 72870.889
Now
The correlation coefficient is 0.9999
b)
No, Since the correlation is greater than 0.997, the calibration cannot be done again.
c)
The equation of regression line is
y = b0 + b1x
where
b1 = slope
b0 = intercept
Formula for finding slope (b1) is
So,
Formula for finding y-intercept (b) is
where
So,
The equation for regression line is
d)
Given
Absorbence, X= 39
The equation for regression line is
So,
Therefore,the estimated nitrate level in a water specimen with absorbence 39 is 328.7 mg/l.
e)
This prediction should be very accurate because the relationship is so strong as indicated by r.