统计学基础实践：第七版

需积分: 10 28 浏览量更新于2024-07-18 收藏 31.46MB PDF 举报

"The Basic Practice of Statistics 7th (DAVID S. MOORE)" 是一本统计学教材，由DAVID S. MOORE、WILLIAM I. NOTZ和MICHAEL A. FLIGNER三位教授合作编写，分别来自普渡大学、俄亥俄州立大学和加利福尼亚大学圣克鲁兹分校。该书基于平衡内容、数据体验和理念重要性的三大原则，这与GAISE（Guidelines in Assessment and Instruction for Statistics Education）项目的学院报告中的主题紧密相连。本书的主要目标是教育和指导学生理解和应用统计学的基本概念和方法。通过平衡内容，它旨在涵盖统计学的各个核心领域，包括描述性统计、概率理论、推断统计、假设检验、回归分析等，确保学生对统计学有全面的认识。体验数据这一原则强调了实践操作的重要性，书中可能包含了大量的实际案例研究、数据集和数据分析活动，让学生能够亲手处理和解析数据，从而深化理解。理念的重要性则意味着书中会探讨统计思想的深层意义，如随机性、不确定性、抽样分布和统计推断的逻辑。这些概念对于培养学生的批判性思维和解决问题的能力至关重要。此外，GAISE报告的指导原则在本书中得到体现，表明其教学方法符合现代统计教育的最佳实践。这本书适合大学本科层次的学生使用，不仅适用于统计学的入门课程，也适用于那些需要应用统计知识的其他学科。通过学习，学生可以掌握如何收集、整理、分析和解释数据，为他们在学术研究或职业生涯中做出基于数据的决策打下坚实基础。该版本为第七版，表明作者团队对教材进行了多次修订和更新，以适应不断发展的统计学领域和教学需求。书中可能包含了最新的统计软件应用、新的数据集和当前领域的实例，使教学内容保持前沿和相关性。此外，提到的“packaged with EESEE/CrunchIt! access card”表明该教材可能随书附带电子资源，如在线练习系统（EESEE）和数据分析工具（CrunchIt!），这些辅助工具旨在增强学生的学习体验，提供即时反馈，帮助他们更好地理解和掌握统计学概念。《The Basic Practice of Statistics》第七版是一本全面、实践导向且与时俱进的统计学教科书，旨在培养学生的统计素养，使他们能够有效地运用统计方法解决实际问题。

Online Resources

Many sections end with references to the most relevant

and helpful online resources (chosen by the authors and

available in LaunchPad) for students to use for further

explanation or practice.

Using Technology

Located where most appropriate, these special sections display and comment on

the output from graphing calculators, spreadsheets, and statistical software in the

context of examples from the text.

2.9 Using technology

Although a calculator with “two-variable statistics” functions will do the basic

calculations we need, more elaborate tools are helpful. Graphing calculators and

computer software will do calculations and make graphs as you command, freeing

you to concentrate on choosing the right methods and interpreting your results.

Figure 2.4 displays output describing the travel times to work of 20 people in New

York State (Example 2.3). Can you find x

, s, and the five-number summary in each

output? The big message of this section is: Once you know what to look for, you can

read output from any technological tool.

The displays in Figure 2.4 come from a Texas Instruments graphing calcula-

tor, the Minitab, CrunchIt!, and JMP statistical programs, and the Microsoft

Excel spreadsheet program. Minitab and JMP allow you to choose what descrip-

tive measures you want, whereas the descriptive measures in the CrunchIt! out-

put are provided by default. Excel and the calculator give some things we don’t

need. Just ignore the extras. Excel’s “Descriptive Statistics” menu item doesn’t

give the quartiles. We used the spreadsheet’s separate quartile function to get

and Q

FREE144_SE_CH02_047-074.indd Page 61 02/09/14 7:52 PM user-f-401 /208/WHF00270/work/indd/CH02

Texas Instruments Graphing Calculator

Minitab

Descriptive Statistics: NYtime

variable Count Mean StDev Variance Minimum Q1 Median Q3 Maximum

NYtime 20 31.25 21.88 478.62 5.00 15.00 22.50 43.75 85.00

Total

FREE144_SE_CH02_047-074.indd Page 62 02/09/14 7:53 PM user-f-401 /208/WHF00270/work/indd/CH02

Microsoft Excel

minutes

Mean

QUARTILE(A2:A21,1)

QUARTILE(A2:A21,3)

31.25

Median

Standard Error

Mode

Standard Deviation

21.8773495

22.5

42.5

4.891924064

Sample Variance

Kurtosis

Skewness

478.6184211

0.329884126

1.040110836

625

Range

Minimum

Maximum

Sum

Count

Sheet4

Sheet1

Sheet

Sheet2

CrunchIt!

JMP Output

Distributions

100% maximum 85

75% quartile 43.75

50% median 22.5

25% quartile 15

0% minimum 5

NYtime

Summary Statistics

Quantiles

Mean 31.25

Std Dev 21.877349

N20

Minutes

Results - Descriptive Statistics

Export

n Sample Mean Standard Deviation Median Q3 MaxMin Q1

31.25 21.88 5 15 22.50 42.50 85

FREE144_SE_CH02_047-074.indd Page 62 02/09/14 7:53 PM user-f-401 /208/WHF00270/work/indd/CH02

FIGURE 2.4

Output from a graphing calculator, three

statistical software packages, and a

spreadsheet program describing the

data on travel times to work in New

York State.

FREE144_SE_CH02_047-074.indd Page 62 02/09/14 7:53 PM user-f-401 /208/WHF00270/work/indd/CH02

Online Resources

• The Snapshots video, Summarizing Quantitative Data, provides an overview

of the need for measures of center and variability as well as some details of

the computations.

• The StatClips Examples video, Summaries of Quantitative Data Example

C, gives the details for the computation of the mean, median, and standard

deviation in a small example. You can verify the computations along with the

video, either by hand or using your technology.

• The StatClips Examples videos, Basic Principles of Exploring Data Example

B and Basic Principles of Exploring Data Example C, emphasize the need to

examine outliers and understand them, rather than simply discarding obser-

vations that don’t seem to fit.

FREE144_SE_CH02_047-074.indd Page 60 02/09/14 7:52 PM user-f-401 /208/WHF00270/work/indd/CH02

FEATURES OF THE BASIC PRACTICE OF STATISTICS ■ Seventh Edition

Statistics in Your World

These brief asides in each chapter illustrate major concepts

or present cautionary tales through entertaining and relevant

stories, allowing students to take a break from the exposition

while staying engaged.

HE SAID,

SHE SAID.

Height, weight, and body mass

distributions in this book come from

actual measurements by a government

survey. That is a good thing. When

asked their weight, almost all women

say they weigh less than they really do.

Heavier men also underreport their

weight—but lighter men claim to

weigh more than the scale shows. We

leave you to ponder the psychology

of the two sexes. Just remember that

“say-so” is no substitute for measuring.

FREE144_SE_CH03_075-100.indd Page 85 02/09/14 5:21 PM user-f-401 /208/WHF00270/work/indd/CH03

Chapter Summary and Link It

Each chapter concludes with a summary of the chapter speciﬁ cs,

including major terms and processes, followed by a brief discus-

sion of how the chapter links to material from both previous and

upcoming chapters.

Link It

In Chapters 1 to 3, we focused on exploring features of a single variable. In this chapter, we

continued our study of exploratory data analysis but for the purpose of examining relation-

ships between variables. A useful tool for exploring the relationship between two variables

is the scatterplot. When the relationship is linear, correlation is a numerical measure of the

strength of the linear relationship.

It is tempting to assume that the patterns we observe in our data hold for values of our

variables that we have not observed—in other words, that additional data would continue

to conform to these patterns. The process of identifying underlying patterns would seem

to assume that this is the case. But is this assumption justified? Parts II to V of the book

answer this

uestion.

FREE144_SE_CH04_101-126.indd Page 118 02/09/14 6:35 PM user-f-401 /208/WHF00270/work/indd/CH04

CHAPTER 4 SUMMARY

hapter Specifics

■

To study relationships between variables, we must measure the variables on the same

group of individuals.

■

If we think that a variable x may explain or even cause changes in another variable y,

we call x an explanatory variable and y a response variable.

■

A scatterplot displays the relationship between two quantitative variables measured on

the same individuals. Mark values of one variable on the horizontal axis (x axis) and

values of the other variable on the vertical axis (y axis). Plot each individual’s data as

a point on the graph. Always plot the explanatory variable, if there is one, on the x axis

of a scatterplot.

■

Plot points with different colors or symbols to see the effect of a categorical variable in

a scatterplot.

■

In examining a scatterplot, look for an overall pattern showing the direction, form, and

strength of the relationship and then for outliers or other deviations from this pattern.

■

Direction: If the relationship has a clear direction, we speak of either positive associa-

tion (high values of the two variables tend to occur together) or negative association

(high values of one variable tend to occur with low values of the other variable).

■

Form: Linear relationships, where the points show a straight-line pattern, are an

important form of relationship between two variables. Curved relationships and clus-

ters are other forms to watch for.

■

Strength: The strength of a relationship is determined by how close the points in the

scatterplot lie to a simple form such as a line.

■

The correlation r measures the direction and strength of the linear association

between two quantitative variables x and y. Although you can calculate a correlation

for any scatterplot, r measures only straight-line relationships.

FREE144_SE_CH04_101-126.indd Page 117 02/09/14 6:35 PM user-f-401 /208/WHF00270/work/indd/CH04

Check Your Skills and

Chapter Exercises

Each chapter ends with a series of multiple-

choice problems that test students’ understand-

ing of basic concepts and their ability to apply

the concepts to real-world statistical situations.

The multiple-choice problems are followed by a

set of more in-depth exercises that allow students

to make judgments and draw conclusions based

on real data and real scenarios.

4.14 Researchers collect data on 5,134 American adults

younger than 60. They measure the reaction times (in

seconds) of each subject to a stimulus on a computer

screen and how many years later the subject died.

The researchers are interested in whether reaction

time can predict time to death (in years). When you

make a scatterplot, the explanatory variable on the x axis

(a) is the reaction time.

(b) is the time to death.

4.15 The researchers in Exercise 4.14 found that people

with slower reaction times tended to die sooner. In

a scatterplot of the reaction time and the number of

years to death, you expect to see

(a) a positive association.

(b) very little association.

4.16 Figure 4.7 is a scatterplot of school GPA against IQ

test scores for 15 seventh-grade students. There is one

low outlier in the plot. The IQ and GPA scores for this

student are

(a) IQ ⫽ 0.5, GPA ⫽ 103.

(b) IQ ⫽ 103, GPA ⫽ 0.5.

4.17 If we leave out the low outlier, the correlation for the

remaining 14 points in Figure 4.7 is closest to

(a) 0.9. (b) ⫺0.9. (c) 0.1.

4.18 What are all the values that a correlation r can pos-

sibly take?

(a) r ⱖ 0 (b) 0 ⱕ r ⱕ 1 (c) ⫺1 ⱕ r ⱕ 1

4.19 If the correlation between two variables is close to 0,

you can conclude that a scatterplot would show

(a) a strong straight-line pattern.

(b) a cloud of points with no visible pattern.

pattern of another form.

4.20 The points on a scatterplot lie very close to a straight

line. The correlation between x and y is close to

(a) ⫺1. (b) 1. (c) either ⫺1 or 1, we can’t say which.

4.21 A statistics professor warns her class that her second

midterm is always harder than the first. She tells her

class that students always score 10 points worse on the

second midterm compared to their score on the first

midterm. This means that the correlation between

students’ scores on the first and second exam is

(a) 1. (b) ⫺1. (c) Can’t tell without seeing the data.

4.22 Researchers asked mothers how much soda (in ounces)

their kids drank in a typical day. They also asked

these mothers to rate how aggressive their kids were

on a scale of 1 to 10, with larger values corresponding

to a greater degree of aggression.

The correlation

between amount of soda consumed and aggression

rating was found to be r ⫽ 0.3. If the researchers had

measured amount of soda consumed in liters instead

CHAPTER 1 EXERCISES

CHECK YOUR SKILLS

FREE144_SE_CH04_101-126.indd Page 119 02/09/14 6:35 PM user-f-401 /208/WHF00270/work/indd/CH04

4.24 Scores at the Masters. The Masters is one of the four

major golf tournaments. Figure 4.8 is a scatterplot of

the scores for the first two rounds of the 2013 Masters

for all the golfers entered. Only the 60 golfers with

the lowest two-round total advance to the final two

rounds (unless several people are tied for 60th place,

in which case all those tied for 60th place advance).

The plot has a grid pattern because golf scores must

be whole numbers.

DATA

MASTR13

(a) Read the graph: What was the lowest score in

the first round of play? How many golfers had this

low score? What were their scores in the second

round?

(b) Read the graph: Alan Dunbar had the highest

score in the second round. What was this score?

What was Dunbar’s score in the first round?

4.25 Happy states. Human happiness or well-being can be

assessed either subjectively or objectively. Subjective

assessment can be accomplished by listening to what

people say. Objective assessment can be made from

data related to well-being such as income, climate,

availability of entertainment, housing prices, lack of

traffic congestion, and so on. Do subjective and ob-

jective assessments agree? To study this, investiga-

tors made both subjective and objective assessments

of happiness for each of the 50 states. The subjective

measurement was the mean score on a life-satisfaction

question found on the Behavioral Risk Factor Surveil-

lance System (BRFSS), which is a state-based system of

health surveys. Lower scores indicate a greater degree

of happiness. To objectively assess happiness, the in-

vestigators computed a mean well-being score (called

h i diff i l ) f h

CHECK YOUR SKILLS

CHAPTER 4 EXERCISES

FREE144_SE_CH04_101-126.indd Page 120 02/09/14 6:35 PM user-f-401 /208/WHF00270/work/indd/CH04

FEATURES OF THE BASIC PRACTICE OF STATISTICS ■ Seventh Edition

xvi

Web Exercises

A ﬁ nal set of exercises asks students to investigate data and statistical issues by

researching topics online. These exercises tend to be more involved and provide

an opportunity for students to dig deep into contemporary issues and special ap-

plications of statistics.

6.34 Promoting women. In academics, faculty typically start as assistant professors,

are promoted to associate professor (and gain tenure), and finally reach the rank of

full professor. Some have argued that women have a harder time gaining promotion

to associate and full professor than do men. Do data support this argument? Search

the web to find the number of faculty by rank and gender at some university. Do

you see a pattern that suggests that the proportion of women decreases as rank

increases? We found several sources of data by doing a Google search on “faculty

head count by rank and gender.” In addition to discussing the pattern you find,

provide the data, the name of the school, and the source of the data.

6.35 Accidental deaths and age. Accidental deaths are shocking and tragic. Do the

ways in which people die by accident change with age? Look at the most recent

Statistical Abstract of the United States (www.census.gov/compendia/

statab/) and make a two-way table that provides the counts of deaths due

to accidents from various causes for three different age groups. What do you

conclude?

6.36 Simpson’s paradox. Find an example of Simpson’s paradox and discuss

how your example illustrates the paradox. Two examples that we found (thanks

to Patricia Humphrey at Georgia Southern University) are www.nytimes.

com/2006/07/15/education/15report.html and online.wsj.com/

article/SB125970744553071829.html.

Exploring the Web

FREE144_SE_CH06_163-178.indd Page 178 05/09/14 9:25 PM user-f-401 /208/WHF00270/work/indd/CH06

1. SAT, ACT, and teacher salaries for 2013 for each of the 50 states and the District of

Columbia are available in the data set SATACT. One could use these data to carry

out analyses for ACT scores similar to those for the SAT scores in Chapters 5 and 6.

For example, repeat the analyses in Exercises 5.50 (page 157) and 5.51 (page 158)

using ACT scores instead of SAT scores.

DATA

SATACT

2. The data set MLB contains hitting, pitching, fielding, salary, and win–loss perfor-

mance data from the 2013 season for all major league baseball teams. These data

can be used to determine the correlation between payroll and winning percent-

age. One can also explore what variables are most highly correlated with winning

percentage, and whether variables that measure pitching performance are more

highly correlated with winning percentage than variables that measure hitting

performance. For example, calculate the correlation between winning percentage

and number of home runs, between winning percentage and batting average, be-

tween winning percentage and ERA, between winning percentage and strikeouts

by pitchers, and between winning percentage and payroll. Which has the highest

correlation? These data are from http://www.baseball-reference.

com/. Visit this website for definitions of several of the variables in the data

set.

DATA

MLB

3. Historical temperature data and whether Punxsutawney Phil saw his shadow are

available in the data set PHIL. Repeat the analysis in Exercise 6.32 (page 177),

but define what constitutes “six more weeks of winter-like weather” differently.

For example, you might decide there were six more weeks of winter-like weather if

average temperatures for March were at least one degree below historical averages.

DATA

PHIL

4. Data from the Ohio Department of Health website are available in the pdf

“2013OHH Detail Tables.” This is a source of many tables that can be used for

further analyses using methods discussed in Chapter 6. For example, conduct an

analysis like that in Exercise 7.52 to investigate the relationship between sex and

strategies about weight (Question 67 in the Tables).

DATA

HEALTH

5. The data set WHAT contains three variables and 3848 observations on each. At

one time, this was considered a large data set and difficult to explore with software.

Use various exploratory methods available in software packages such as JMP and

Minitab to find the “hidden pattern” in these data.

DATA

WHAT

Online Data for Additional Analyses

FREE144_SE_CH07_179-200.indd Page 200 02/09/14 8:43 PM user-f-401 /208/WHF00270/work/indd/CH07

Online Data for Additional Analyses

References to larger data sets are suggested in Chapter 7 to provide an opportunity for

students to apply the methods of Chapters 1–6 to explore data on their own. This is

intended to reinforce the idea of exploratory data analysis as a tool for exploring data.

FEATURES OF THE BASIC PRACTICE OF STATISTICS ■ Seventh Edition

xvii

Why Did You Do That?

There is no single best way to organize our presentation of statistics to beginners.

That said, our choices reﬂ ect thinking about both content and pedagogy. Here are

comments on several “frequently asked questions” about the order and selection of

material in The Basic Practice of Statistics.

• Why does the distinction between population and sample not appear in Part I?

There is more to statistics than inference. In fact, statistical inference is appropriate

only in rather special circumstances. The chapters in Part I present tools and tactics

for describing data—any data. These tools and tactics do not depend on the idea of

inference from sample to population. Many data sets in these chapters (for example,

the several sets of data about the 50 states) do not lend themselves to inference be-

cause they represent an entire population. John Tukey of Bell Labs and Princeton, the

philosopher of modern data analysis, insisted that the population–sample distinction

be avoided when it is not relevant. He used the word “batch” for data sets in general.

We see no need for a special word, but we think Tukey was right.

• Why not begin with data production? We prefer to begin with data exploration

(Part I), as most students will use statistics mainly in settings other than planned

research studies in their future employment. We place the design of data produc-

tion (Part II) after data analysis to emphasize that data-analytic techniques apply

to any data. However, it is equally reasonable to begin with data production—the

natural ﬂ ow of a planned study is from design to data analysis to inference. Because

instructors have strong and differing opinions on this question, these two topics are

now the ﬁ rst two parts of the book, with the text written so that it may be started

with either Part I or Part II while maintaining the continuity of the material.

• Why do Normal distributions appear in Part I? Density curves such as the Nor-

mal curves are just another tool to describe the distribution of a quantitative variable,

along with stemplots, histograms, and boxplots. Professional statistical software offers

to make density curves from data just as it offers histograms. We prefer not to sug-

gest that this material is essentially tied to probability, as the traditional order does.

And we ﬁ nd it helpful to break up the indigestible lump of probability that troubles

students so much. Meeting Normal distributions early does this and strengthens the

“probability distributions are like data distributions” way of approaching probability.

• Why not delay correlation and regression until late in the course, as was

traditional? The Basic Practice of Statistics begins by offering experience working

with data and gives a conceptual structure for this nonmathematical but essential

part of statistics. Students proﬁ t from more experience with data and from seeing

the conceptual structure worked out in relations among variables as well as in

describing single-variable data. Correlation and least-squares regression are very

important descriptive tools and are often used in settings where there is no popula-

tion–sample distinction, such as studies of all of a ﬁ rm’s employees. Perhaps most

important, the approach taken by The Basic Practice of Statistics asks students to

think about what kind of relationship lies behind the data (confounding, lurking

variables, association doesn’t imply causation, and so on), without overwhelming

them with the demands of formal inference methods. Inference in the correlation

and regression setting is a bit complex, demands software, and often comes right at

the end of the course. We ﬁ nd that delaying all mention of correlation and regres-

sion to that point means that students often don’t master the basic uses and proper-

ties of these methods. We consider Chapters 4 and 5 (correlation and regression)

essential and Chapter 26 (regression inference) optional.

FEATURES OF THE BASIC PRACTICE OF STATISTICS ■ Seventh Edition

xviii

• Why use the

procedures for a population mean to introduce the reasoning

of inference? This is a pedagogical issue, not a question of statistics in practice.

The two most popular choices for introducing inference are z for a mean and z

for a proportion. (Another option is resampling and permutation tests. We have

included material on these topics, but have not used them to introduce inference.)

We ﬁ nd z for means quite accessible to students. Positively, we can say up front

that we are going to explore the reasoning of inference in the overly simple setting

described in the box on page 374 titled Simple Conditions for Inference about a

Mean. As this box suggests, exactly Normal population and true simple random

sample are as unrealistic as known ␴. All the issues of practice—robustness against

lack of Normality and application when the data aren’t an SRS as well as the need

to estimate ␴—are put off until, with the reasoning in hand, we discuss the practi-

cally useful t procedures. This separation of initial reasoning from messier practice

works well.

Negatively, starting with inference for p introduces many side issues: no exact

Normal sampling distribution, but a Normal approximation to a discrete distribu-

tion; use of p

in both the numerator and denominator of the test statistic to esti-

mate both the parameter p and p

’s own standard deviation; loss of the direct link

between test and conﬁ dence interval; and the need to avoid small and moderate

sample sizes because the Normal approximation for the test is quite unreliable.

There are advantages to starting with inference for p. Starting with z for means

takes a fair amount of time and the ideas need to be rehashed with the introduc-

tion of the t procedures. Many instructors face pressure from client departments

to cover a large amount of material in a single semester. Eliminating coverage of

the “unrealistic” z for means with known variance enables instructors to cover

additional, more realistic applications of inference. Also, many instructors believe

that proportions are simpler and more familiar to students than means. For instruc-

tors who would prefer to introduce inference with z for a proportion, we recom-

mend our book, Statistics in Practice.

• Why didn’t you cover Topic X? Introductory texts ought not to be encyclope-

dic. We chose topics on two grounds: they are the most commonly used in practice,

and they are suitable vehicles for learning broader statistical ideas. Students who

have completed the core of the book, Chapters 1 to 12 and 15 to 24, will have little

difﬁ culty moving on to more elaborate methods. Chapters 25 to 27 offer a choice of

slightly more advanced topics, as do the four companion chapters available online.

xix

FEATURES OF THE BASIC PRACTICE OF STATISTICS ■ Seventh Edition

剩余799页未读，继续阅读

l1101643072

粉丝: 0
资源: 1

统计学基础实践：第七版

The practice of statistics for AP exam (5版)

Introduction to the practice of statistics

皮尔逊相关系数参考文献

关于影响因素分析的参考文献

多元回归模型的参考文献

shape mismatch: value array of shape (2122,8) could not be broadcast to indexing result of shape(2122,7)

C语言与循环有关的英语文献并且以英文参考文献引用格式发给我

接下来请你用英文回答以下问题：“Who is Xuan Zang?Tell his story in your own words."

数字图像处理有哪些参考文献

SQL SERVER INSTR

最新资源