# engineering statistics montgomery 5th edition pdf free

Interested in making videos for your class projects or lectures? Just need a basic editor to be able…. Do you use online videos in your lessons or presentations?

As of today we have 83,, eBooks for you to download for free. State any limitations or assumptions of the model. Conduct appropriate experiments and collect data to test or validate the tentative model or conclusions made in steps 2 and 3. Manipulate the model to assist in developing a solution to the problem. Draw conclusions or make recommendations based on the problem solution. Develop Identify Propose or Manipulate Confirm Draw a clear the important refine the model the solution conclusions description a model and make of the problem factors recommendations Collect data Figure The engineering problem-solving method.

Steps 2—4 in Fig. Many of the engineering Many aspects of engineering practice involve collecting, working with, and using data in the sciences are employed in solution of a problem, so knowledge of statistics is just as important to the engineer as the engineering problem- knowledge of any of the other engineering sciences. For example, consider the gasoline mileage performance of your car.

Do you always get as thermodynamics and exactly the same mileage performance on every tank of fuel? Of course not—in fact, sometimes heat transfer the mileage performance varies considerably. These factors represent potential sources of variability in the system. Statistics gives us a framework for describing this variability and for learning about which potential sources of variability are the most important or have the greatest impact on the gasoline mileage performance.

We also encounter variability in most types of engineering problems. For example, suppose that an engineer is developing a rubber compound for use in O-rings. The O-rings are to be employed as seals in plasma etching tools used in the semiconductor industry, so their resistance to acids and other corrosive substances is an important characteristic.

The tensile strengths in psi of the eight O-rings are , , , , , , , and As we should have anticipated, not all the O-ring specimens exhibit the same measurement of tensile strength. There is variability in the tensile strength measurements. Because the measurements exhibit variability, we say. The constant remains uncover patterns in data. However, this never happens in engineering practice, so the actual measurements we observe exhibit variability.

Figure is a dot diagram of the O-ring tensile strength data. The dot diagram is a very useful plot for displaying a small body of data, say, up to about 20 observations. This plot allows us to easily see two important features of the data: the location, or the middle, and the scatter or variability. The need for statistical thinking arises often in the solution of engineering problems. Consider the engineer developing the rubber O-ring material. From testing the initial specimens, he knows that the average tensile strength is Eight O-ring specimens are made from this modified rubber compound and subjected to the nitric acid emersion test described earlier.

The tensile test results are , , , , , , , and The tensile test data for both groups of O-rings are plotted as a dot diagram in Fig.

However, there are some obvious questions to ask. For instance, how do we know that another set of O-ring specimens will not give different results?

In other words, are these results due entirely to chance? Is a sample of eight O-rings adequate to give reliable results? Statistical inference is has no effect on tensile strength? Statistical thinking and methodology can help answer the process of deciding if these questions. This reasoning is from a sample such as the eight rubber O-rings to a population such as the O-rings that will be sold to customers and is re- ferred to as statistical inference.

See Fig. Clearly, reasoning based on measurements from some objects to measurements on all objects can result in errors called sampling errors. We can think of each sample of eight O-rings as a random and representative sample of all parts that will ultimately be manufactured. The order in which each O-ring was tested was also randomly determined. This is an example of a completely randomized designed experiment.

Sometimes the objects to be used in the comparison are not assigned at random to the treatments. For example, the September issue of Circulation a medical journal pub- lished by the American Heart Association reports a study linking high iron levels in the body with increased risk of heart attack.

The researchers just tracked the subjects over time. This type of study is called an observational study. Designed experiments and observational studies are discussed in more detail in the next section. For example, the difference in heart attack risk could be attributable to the dif- ference in iron levels or to other underlying factors that form a reasonable explanation for the observed results—such as cholesterol levels or hypertension.

In the engineering environment, the data are almost always a sample that has been se- lected from some population. In the previous section we introduced some simple methods for summarizing and visualizing data.

In the engineering environment, the data are almost always a sample that has been selected from some population. A sample is a subset of the population containing the observed objects or the outcomes and the resulting data. When little thought is put into the data collection procedure, serious problems for both the statistical analysis and the practical inter- pretation of results can occur.

Montgomery, Peck, and Vining describe an acetone-butyl alcohol distillation col- umn. A schematic of this binary column is shown in Fig. The production personnel very infrequently change this rate. In most such studies, the engineer is interested in using the data to construct a model relating the variables of interest.

These types of models are called empirical models, and they are illustrated in more detail in Section A retrospective study takes advantage of previously collected, or historical, data. It has the advantage of minimizing the cost of collecting the data for the study. However, there are several potential problems: 1. The historical data on the two temperatures and the acetone concentration do not correspond directly.

Constructing an approximate correspondence would probably require making several assumptions and a great deal of effort, and it might be impos- sible to do reliably. Because the two temperatures do not vary. Within the narrow ranges that they do vary, the condensate temperature tends to in- crease with the reboil temperature. Retrospective studies, although often the quickest and easiest way to collect engineering process data, often provide limited useful information for controlling and analyzing a process.

In general, their primary disadvantages are as follows: 1. Some of the important process data often are missing. The reliability and validity of the process data are often questionable. The nature of the process data often may not allow us to address the problem at hand. The engineer often wants to use the process data in ways that they were never in- tended to be used.

Using historical data always involves the risk that, for whatever reason, some of the important data were not collected or were lost or were inaccurately transcribed or recorded. Consequently, historical data often suffer from problems with data quality.

These errors also make historical data prone to outliers. Just because data are convenient to collect does not mean that these data are useful. Historical data cannot provide this information if information on some important variables was never collected.

For example, the ambient temperature may affect the heat losses from the distillation column. On cold days, the column loses more heat to the environment than during very warm days.

The production logs for this acetone-butyl alcohol column do not routinely record the ambient temperature.

Also, the concentration of acetone in the input feed stream has an effect on the acetone concentra- tion in the output product stream. However, this variable is not easy to measure routinely, so it is not recorded either. Consequently, the historical data do not allow the engineer to include either of these factors in the analysis even though potentially they may be important.

The purpose of many engineering data analysis efforts is to isolate the root causes underly- ing interesting phenomena. With historical data, these interesting phenomena may have occurred months, weeks, or even years earlier. Analyses based on historical data often identify interesting phenomena that go unexplained. Finally, retrospective studies often involve very large indeed, even massive data sets.

As the name implies, an observational study simply observes the process or population during a period of routine operation. Usually, the engineer interacts or disturbs the process only as much as is required to obtain data on the system, and often a special effort is made to collect data on variables that are not routinely recorded, if it is thought that such data might be useful.

With proper planning, observational studies can ensure accurate, complete, and reliable data. The data col- lection form should provide the ability to add comments to record any other interesting phenomena that may occur, such as changes in ambient temperature. It may even be possible to arrange for the input feed stream acetone concentration to be measured along with the other variables during this relatively short-term study.

An observational study conducted in this manner would help ensure accurate and reliable data collection and would take care of problem 2 and possibly some aspects of problem 1 associated with the retrospective study.

This approach also minimizes the chances of observing an outlier related to some error in the data. Unfortunately, an observational study cannot address problems 3 and 4. Observational studies can also involve very large data sets. In a designed experiment, the engineer makes deliberate or purposeful changes in controllable variables called factors of the system, observes the resulting system output, and then makes a decision or an inference about which variables are responsible for the changes that he or she observes in the output performance.

An important distinction between a designed experiment and either an observational or retrospective study is that the different combinations of the factors of interest are applied randomly to a set of experimental units.

Montgomery Synopsis: Montgomery, Runger, and Hubele provide modern coverage of engineering statistics, focusing on how statistical tools are integrated into the engineering problem-solving process. All major aspects of engineering statistics are covered, including descriptive statistics, probability and probability distributions, statistical test and confidence intervals for one and two samples, building regression models, designing and analyzing engineering experiments, and statistical process control.

Developed with sponsorship from the National Science Foundation, this revision incorporates many insights from the authors teaching experience along with feedback from numerous adopters of previous editions.

As noted above, this approximate procedure will be valid as long as p is not extremely close to 0 or 1, and if the sample size is relatively large. The following result will be used to perform hypothesis testing and to construct confidence intervals on p. Let X be the number of observations in a random sample of size n that belongs to the class associated with p.

Because the test statistic follows a standard normal distribution if H0 is true, the P-value is calculated exactly like the P-value for the z-tests in Section The customer requires that the process fallout or fraction defective at a critical manufacturing step not exceed 0.

The semiconductor manufacturer takes a random sample of devices and finds that 4 of them are defective. Can the manufacturer demonstrate process capability for the customer?

We may solve this problem using the seven-step hypothesis testing procedure as follows: 1. Parameter of interest: The parameter of interest is the process fraction defective p. The practical engineering conclusion is that the process is capable.

The following Minitab output shows the results for Example In Section This Minitab display shows the result of using the normal approximation for tests and CIs. When the sample size is small, this may be inappropriate. Small Sample Tests on a Binomial Proportion Tests on a proportion when the sample size n is small are based on the binomial distribution, not the normal approximation to the binomial.

To illustrate, suppose we wish to test H0: p 6 p0. Let X be the number of successes in the sample. The P-value for this test would be found from the lower tail of a binomial distribution with parameters n and p0. Specifically, the P-value would be the probability that a binomial random variable with parameters n and p0 is less than or equal to X.

P-values for the upper-tail one-sided test and the two-sided alternative are computed similarly. Minitab will calculate the exact P-value for a binomial test. The output below contains the exact P-value results for Example Notice that the CI is different from the one found using the normal approximation. Suppose that p is the true value of the population proportion.

The sample size equation follows. Unfortunately, the upper and lower limits of the CI obtained from this equation contain the unknown parameter p. This procedure depends on the adequacy of the normal approximation to the binomial. In situations where this approximation is inappropriate, particularly in cases where n is small, other methods must be used.

One approach is to use tables of the binomial distribution to obtain a confidence interval for p. However, we prefer to use numerical methods based on the binomial probability mass function that are implemented in computer programs. This method is used in Minitab and is illustrated for the situation of Example in the boxed display on page EXAMPLE Crankshaft Bearings In a random sample of 85 automobile engine crankshaft bearings, 10 have a surface finish that is rougher than the specifications allow.

A Different Confidence Interval on a Binomial Proportion There is a different way to construct a CI on a binomial proportion than the traditional approach in equation Was that appropriate? How can you do this without performing any additional calculations?

Consider the following Minitab output. Of randomly selected cases of lung cancer, resulted in death. Large passenger vans are thought to have a high propensity of rollover accidents when fully loaded.

Thirty accidents of these vans were examined, and 11 vans had rolled over. Use an initial estimate of p from this problem. A random sample of 50 suspension helmets used by motorcycle riders and automobile race-car drivers was subjected to an impact test, and on 18 of these helmets some damage was observed. Explain how this confidence interval can be used to test the hypothesis in part a. The Arizona Department of Transportation wishes to survey state residents to determine what proportion of the population would be in favor of building a citywide light-rail c04DecisionMakingforaSingleSample.

A manufacturer of electronic calculators is interested in estimating the fraction of defective units produced. A random sample of calculators contains 10 defectives. A study is to be conducted of the percentage of homeowners who have a high-speed Internet connection.

The fraction of defective integrated circuits produced in a photolithography process is being studied. A random sample of circuits is tested, revealing 18 defectives. Consider the defective circuit data and hypotheses in Exercise An article in Fortune September 21, claimed that one-half of all engineers continue academic studies beyond the B.

Data from an article in Engineering Horizons Spring indicated that of new engineering graduates were planning graduate study. A manufacturer of interocular lenses is qualifying a new grinding machine. A random sample of lenses contains 11 defective lenses. A sample of helmets revealed that 24 helmets contained such defects. A random sample of registered voters in Phoenix is asked whether they favor the use of oxygenated fuels year round to reduce air pollution. The warranty for batteries for mobile phones is set at operating hours, with proper charging procedures.

A study of batteries is carried out and three stop operating prior to hours. Do these experimental results support the claim that less than 0. Consider the lung cancer data given in Exercise Consider the helmet data given in Exercise Consider the knee surgery data given in Exercise We may also want to find a range of likely values for the variable associated with making the prediction.

This is a different problem than estimating the mean of that random variable, so a CI on the mean is not really appropriate. Suppose that you plan to purchase a new driver of the type that was tested in that example. What is a reasonable prediction of the coefficient of restitution for the driver that you purchase which is not one of the clubs that was tested in the study , and what is a range of likely values for the coefficient of restitution?

Suppose that X1, X2,. The prediction error is normally distributed because the original observations are normally distributed. Finally, recall that CIs and hypothesis tests on the mean are relatively insensitive to the normality assumption. PIs, on the other hand, do not share this nice feature and are rather sensitive to the normality assumption because they are associated with a single future value drawn at random from the normal distribution. We plan to buy a new golf club of the type tested.

What is a likely range of values for the coefficient of restitution for the new club? The normal probability plot in Fig. A reasonable point prediction of its coefficient of restitution is the sample mean, 0.

Notice that the prediction interval is considerably longer than the CI on the mean. Then the interval from 0. However, because of sampling variability in x and s, it is likely that this interval will contain c04DecisionMakingforaSingleSample.

The solution is to replace 1. Fortunately, it is easy to do this. One-sided tolerance bounds can also be computed. Consider the tire life problem described in Exercise Consider the Izod impact strength problem described in Exercise Consider the life of biomedical devices described in Exercise Consider the fatty acid content of margarine described in Exercise Consider the breakdown voltage of diodes described in Exercise Consider the metal rods described in Exercise Another kind of hypothesis is often encountered: We do not know the underlying distribution of the population, and we wish to test the hypothesis that a particular distribution will be satisfactory as a population model.

For example, we might wish to test the hypothesis that the population is normal. In Chapter 3, we discussed a very useful graphical technique for this problem called probability plotting and illustrated how it was applied in the case of normal, lognormal, and Weibull distributions. In this section, we describe a formal goodness-of-fit test procedure based on the chi-square distribution. The test procedure requires a random sample of size n from the population whose probability distribution is unknown.

These n observations are arranged in a histogram, having k bins or class intervals. Let Oi be the observed frequency in the ith class interval. From the hypothesized probability distribution, we compute the expected frequency in the ith class interval, denoted Ei. This approximation improves as n increases.

We would reject the hypothesis that the distribution of the population is the hypothesized distribution if the calculated value of the test statistic X 20 is too large. That c04DecisionMakingforaSingleSample. One point to be noted in the application of this test procedure concerns the magnitude of the expected frequencies. If these expected frequencies are too small, the test statistic X20 will not reflect the departure of observed from expected, but only the small magnitude of the expected frequencies.

There is no general agreement regarding the minimum value of expected frequencies, but values of 3, 4, and 5 are widely used as minimal. Some writers suggest that an expected frequency could be as small as 1 or 2, so long as most of them exceed 5. Should an expected frequency be too small, it can be combined with the expected frequency in an adjacent class interval. The corresponding observed frequencies would then also be combined, and k would be reduced by 1. Class intervals are not required to be of equal width.

We now give an example of the test procedure. The following data result: Number of Defects Observed Frequency 0 1 2 3 32 15 9 4 Is it reasonable to conclude that the number of defects is Poisson distributed?

The mean of the assumed Poisson distribution in this example is unknown and must be estimated from the sample data. From the Poisson distribution with parameter 0. The expected frequencies are shown next. Number of Defects Probability Expected Frequency 0 1 2 3 or more 0.

Parameter of interest: The parameter of interest is the form of the distribution of defects in printed circuit boards. Null hypothesis, H0: The form of the distribution of defects is Poisson. Alternative hypothesis, H1: The form of the distribution of defects is not Poisson. Reject H0 if: Reject H0 if the P-value is less than 0. Therefore, since the P-value is greater than 0. This value was computed using Minitab.

Consider the following frequency table of observations on the random variable X. Values 0 1 2 3 4 5 Observed Frequency 8 25 23 21 16 7 a Based on these observations, is a Poisson distribution with a mean of 2.

Let X denote the number of flaws observed on a large coil of galvanized steel. Seventy-five coils are inspected, and the following data were observed for the values of X. Values 1 2 3 4 5 6 7 8 Observed Frequency 1 11 8 13 11 12 10 9 a Does the assumption of a Poisson distribution with a mean of 6. The number of calls arriving to a switchboard from noon to 1 P. Let X be defined as the number of calls during that 1-hour period.

The observed frequency of calls was recorded and reported as follows: Value 5 7 8 9 10 Observed Frequency Value 4 11 4 12 4 13 5 14 1 15 Observed Frequency 3 3 1 4 1 c04DecisionMakingforaSingleSample. The number of cars passing eastbound through the intersection of Mill Avenue and University Avenue has been tabulated by a group of civil engineering students.

They have obtained the following data: Vehicles per Minute 40 41 42 43 44 45 46 47 48 49 50 51 52 Page Observed Frequency Vehicles per Minute Observed Frequency 14 24 57 53 54 55 56 57 58 59 60 61 62 63 64 65 96 90 81 73 64 61 59 50 42 29 18 15 a Does the assumption of a Poisson distribution seem appropriate as a probability model for this process? Define X as the number of underfilled bottles in a filling operation in a carton of 12 bottles. Eighty cartons are inspected, and the following observations on X are recorded.

Kenyon alumni working in food and drink industries are sharing their go-to happy hour pairings. Interested in making videos for your class projects or lectures? Just need a basic editor to be able….