Scribd
Upload
56 views

Engineering Data Solutions

1) This document provides information on various statistical concepts including measures of central tendency, variance and standard deviation, the standard normal distribution, the normal distribution, confidence limits, and linear regression. 2) Key concepts covered include finding the median, mode, and mean of data sets. Calculating variance and standard deviation for samples of data. Using the standard normal distribution to find probabilities of z-scores. Finding z-scores and using the normal distribution table. Calculating confidence intervals for means. Determining the line of best fit and correlation for linear regression models. 3) Examples are provided to demonstrate calculating and interpreting these statistical metrics for given data sets. Concepts are explained through worked examples to find measures of central tendency

Uploaded by

ppp2010
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
56 views

Engineering Data Solutions

1) This document provides information on various statistical concepts including measures of central tendency, variance and standard deviation, the standard normal distribution, the normal distribution, confidence limits, and linear regression. 2) Key concepts covered include finding the median, mode, and mean of data sets. Calculating variance and standard deviation for samples of data. Using the standard normal distribution to find probabilities of z-scores. Finding z-scores and using the normal distribution table. Calculating confidence intervals for means. Determining the line of best fit and correlation for linear regression models. 3) Examples are provided to demonstrate calculating and interpreting these statistical metrics for given data sets. Concepts are explained through worked examples to find measures of central tendency

Uploaded by

ppp2010
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Discrete vs.

Continuous Solutions
1) a) b) c) d) Continuous Discrete Discrete Continuous

Measures of Central Tendency


2) 3)

In order to find the median, first rank the data: 42 67 59 68 60 68 61 70 61 71 61 74 65 74 67 78

Now find the location of the median:

A fraction appears due to the fact that there is an even number of data values. Therefore, the median is the mean of the two middle numbers i.e. 67 and 67 in this case. Clearly, the answer is 67. 4) a) No repetition, therefore there is no mode. b) Two modes (bimodal) since 7 and 3 both appear twice. c) The mode is 0.

Variance and Deviation


5) We first need to find the sample mean:

Now, a table is helpful:

x 6.1 5.7 5.8 6 5.8 6.3 0.15 -0.25 -0.15 0.05 -0.15 0.35 0.0225 0.0625 0.0225 0.0025 0.0225 0.1225 Sum = 0.255

The sample variance is:

Therefore, the sample standard deviation is:

Since s < 0.3, the production process does not need assessment. 6) The data represents a sample since they only deal with 6 years of the highways lifespan. We first need the sample mean:

Table:
x 24 16 27 34 36 25

-3 -11 0 7 9 -2

9 121 0 49 81 4 Sum = 264

Sample Variance:

Sample Standard Deviation:

Standard Normal Distribution


7) We are dealing with standard normal variables, and thus the distribution has a mean of 0 and a standard deviation of 1. The standard normal table may be used directly. a) P(z < 0) = 0.5, since z = 0 represents the exact middle of the curve and the total area is equal to 1. b) P(-.67 < z < 0) = .2486, the same area as 0 to +.67 since symmetric. c) P(-2.3 < z < -1.45) = P(0 < z < 2.3) - P(0 < z < 1.45) = .4893 - .4265 = .0628 d) P(-.73 < z < 2.31) = P(0 < z < .73) + P(0 < z < 2.31) = .2673 + .4896 = .7569 e) P(z < 1.96) = P(z < 0) + P(0 < z < 1.96) = 0.5 + .4750 = .9750 f) P(-! < z < !) = P(-1 < z < 1) = .3413 + .3413 = .6826 g) P(-3! < z < 3!) = P(-3 < z < 3) = .4987 + .4987 = .9974

Normal Distribution
8) In this case, we need to find the z score in order to use the standard normal table:

a) x = 46 , therefore z = 0. The area for which z > 0 is .5, meaning the percentage is 50. b) x = 50, therefore z = 1. The area for which z > 1 is .5 - .3413, meaning the percentage is 15.87. c) x = 49, therefore z = .75. The area for which z < .75 is .5 + .2734, meaning the percentage is 77.34. d) x = 45 and 49, therefore z = -.25 and .75. The area for which -.25 < z < .75 is .0987 + .2734, meaning the percentage is 37.21. e) x = 56 and 46, therefore z = 2.5 and 0. The area for which z > 2.5 or z < 0 is (.5 - .4938) + .5 = .5062, meaning a percentage of 50.62. f) Within 1.5 standard deviations of the mean corresponds to 1.5*4 = +/- 6 units from the mean of 46. Therefore, x = 40 and 52 and z = -1.5 and 1.5. This shows that using the z score method is really only changing the working units over to

standard deviations. The area for which 1.5 < z < 1.5 is 2*(.4332), meaning a percentage of 86.64. g) Outside 2.3 standard deviations corresponds to 2.3 > z > 2.3 and an area of 2*(.5-.4893). This gives a percentage of 2.14. 9) = 48 months and ! = 9 months Need to determine the expression for the z score:

a) We want the number of sensors that will last less than 3 years, or 36 months. This means we are interested in the region to the left of 36 months on the normal curve. When x = 36, z = -1.33 The area for which z < -1.33 is (.5 - .4082) = .0918 or 9.18% Therefore, 9.18% of the sensors will not satisfy the guarantee. b) 1% corresponds to an area of .01 under the curve. The following figure will aid in this problem:

Replace

Since the tails of the curve go off into infinity, we must work from the mean instead of the extreme left. Subtract .01 from .5 in order to get the desired area. This value is .49. We can now find the z value for an area of .49 from the standard normal table. It turns out that z = -2.33 is pretty close (the same as z = 2.33). Knowing the z value, we can determine the length of the guarantee by switching the units through the use of the z score:

Therefore, a guarantee of 27 months will ensure that only 1% will need to be replaced.

Confidence Limits
10) First step is to decide which category the sample size falls under, small or large. a) Large sample, " = .025, z.025 = 1.96 from table.

Therefore, 95% of the time 121.08 < < 128.92. b) Large sample, " = .005, z.005 = 2.575 from table.

Therefore, 99% of the time 199.55 < < 212.45. c) Small sample, " = .025, df = 28, t.025 = 2.048 from table.

Therefore, 95% of the time 106.44 < < 115.56. d) Large sample, " = .05, z.05 = 1.65 from table.

Therefore, 90% of the time 153.45 < < 154.55. e) Small sample, " = .005, df = 14, t.005 = 2.977 from table.

Therefore, 99% of the time 49.93 < < 56.07. 11) Assume that the number of bolts will be a large sample. The maximum error on the sample mean is to be .001 inch, therefore:

The confidence is 90%, thus " = .05. From the table, z.05 = 1.65, s didnt change. From the above equation:

Therefore, 273 bolts will need to be measured in order for the sample mean to be within .001 inch of the mean 90% of the time.

Linear Regression and Correlation


12) a)
x 2 2 4 4 6 6 24 y 9 7 5 5 3 1 30 112 96 4 5 x^2 4 4 16 16 36 36 x*y 18 14 20 20 18 6

Sum x Sum y Sum x^2 Sum x*y Mean x Mean y

Therefore,

is the line of best fit.

b)
x 1 2 4 5 12 y 2 6 14 18 40 46 160 3 10 x^2 1 4 16 25 x*y 2 12 56 90

Sum x Sum y Sum x^2 Sum x*y Mean x Mean y

Therefore, 13) a)
x 0 0 0 1 2 2 2 7 y 2 4 6 6 7 8 9 42

is the line of best fit.

x*y 0 0 0 6 14 16 18

x^2 0 0 0 1 4 4 4

y^2 4 16 36 36 49 64 81

Sum x Sum y Sum x*y Sum x^2 Sum y^2

54 13 286

Therefore, there is a reasonable linear relationship with a positive slope.

b)
x 1 2 3 4 5 15 y 66 42 32 13 21 174 403 55 7754 x*y 66 84 96 52 105 x^2 1 4 9 16 25 y^2 4356 1764 1024 169 441

Sum x Sum y Sum x*y Sum x^2 Sum y^2

Therefore, there is a strong linear relationship with a negative slope.

You might also like