15-1 Projects

Key Concept A final project is an excellent activity that can be a valuable and re-
warding experience for students in the introductory statistics course. This final
project provides students with the opportunity to use principles of statistics in a real
and interesting application. This section provides a suggested format for such a project.
Group Project vs. Individual Project Although different topics could be as-
signed to individuals, group projects tend to be more effective because they help de-
velop the interpersonal skills that are so necessary in today’s working environment.
One study showed that the “inability to get along with others” is the main reason for
firing employees, so a group project can be very helpful in preparing students for
their future work environments. Groups of three, four, or five students work well.
The professor should select groups with consideration given to important factors
such as past class performance and attendance.

Oral Report A 10- to 15-minute class presentation should involve all group mem-
bers in a coordinated effort to clearly describe the important components of the
study. Students typically have some reluctance to speak in public, so a brief oral re-
port can be very helpful in building confidence. The oral report is an activity that can
better prepare students for future professional activities.

Written Report The main objective of the project is not to produce a written
document equivalent to a term paper. However, a brief written report should be sub-
mitted, and it should include the following components:
1. List of data collected along with a description of how the data were obtained.
2. Description of the method of analysis
3. Relevant graphs and>or statistics, including STATDISK, Minitab, Excel, or
TI-83>84 Plus displays
4. Statement of conclusions
5. Reasons why the results might not be correct, along with a description of ways
in which the study could be improved, given sufficient time and money
Large Classes or Online Classes: Posters or PowerPoint Some classes are
too large for individual or group projects. Online classes are not able to meet as a
group. For such classes, reports of individual or small group projects can be presented

742
 15-1 Projects 743

through posters similar to those found at conference poster sessions. Posters or Power-
Point presentations summarizing important elements of a project can be submitted to
professors for evaluation.

Project Topics The “Cooperative Group Activities” listed near the end of each
chapter include more than 100 suggestions for projects. The following comments
about a survey can be another excellent source of project topics.

Survey A survey can be an excellent source of data that can be used in a statistics
project. The sample survey below collects information that can be used to address
questions such as these:
1. When people “randomly” select digits (as in Question 2), are the results actu-
ally random?
2. Do the last four digits of social security numbers appear to be random?
3. Do males and females carry different amounts of change?
4. Do males and females have different numbers of credit cards?
5. Is there a difference in pulse rates between those who exercise and those who do
not?
6. Is there a difference in pulse rates between those who smoke and those who do
not?
7. Is there a relationship between exercise and smoking?
8. Is there a relationship between eye color and exercise?
9. Is there a relationship between exercise and the number of hours worked each
week?
10. Is there a correlation between height and pulse rate?

Survey

1. _______ Female _______ Male 9. Record your pulse rate by counting the number of
heartbeats for 1 minute: __________________
2. Randomly select four digits and enter them here:
_______ _______ _______ _______ 10. Doyou exercise vigorously (such as running, swim-
ming, cycling, tennis, basketball, etc.) for at least 20
3. Eye color: __________________
minutes at least twice a week? _______ Yes _______ No
4. Enter your height in inches: __________________
11. How many credit hours of courses are you taking this
5. What is the total value of all coins now in your semester? __________________
possession? __________________
12. Are you currently employed? _______ Yes _______ No
6. How many keys are in your possession at this time? If yes, how many hours do you work each week?
__________________ __________________
7. How many credit cards are in your possession at this 13. During the past 12 months, have you been the driver
time? _______ of a car that was involved in an accident?
_______Yes _______ No
8. Enter the last four digits of your social security num-
ber. (For reasons of security, you can rearrange the or- 14. Do you smoke? _______ Yes _______ No
der of these digits.) _______ _______ _______
_______ _______ 15. _______Left-handed _______ Right-handed
_______ Ambidextrous
744 Chapter 15 Projects, Procedures, Perspectives

15-2 Procedures
Key Concept This section describes a general approach for the statistical analysis of data.
Context, Source, Sampling Method Instead of mindlessly plugging data into
some particular statistical procedure, we should begin with some basic considera-
tions, including these:
1. Clearly identify the context of the data (as discussed in Section 1-2).

2. Consider the source of the data and determine whether that source presents any
issues of bias that might affect the validity of the data (as discussed in Section 1-2).
3. Consider the sampling method to ensure that it is the type of sampling likely to
result in data that are representative of the population. Be especially wary of
voluntary response samples.
Exploring, Comparing, Describing After collecting data, first consider exploring,
describing, and comparing data. Start with an appropriate graph for the data. (For ex-
ample, with sample data consisting of single values, construct a histogram, normal
quantile plot, and boxplot. With paired data, construct a scatterplot.) Using the basic
tools described in Chapters 2 and 3, consider the following:
1. Center: Find the mean and median, which are measures of center giving us an
indication of where the middle of the data set is located.
2. Variation: Find the range and standard deviation, which are measures of the
amount that the sample values vary among themselves.
3. Distribution: Construct a histogram to see the shape of the distribution of the
data. Also construct a normal quantile plot and determine if the data are from
a population having a normal distribution.
4. Outliers: Identify any sample values that lie very far away from the vast majority
of the other sample values. If there are outliers, try to determine whether they
are errors that should be corrected. If the outliers are correct values, study their
effects by repeating the analysis with the outliers excluded.
5. Time: Determine if the population is stable or if its characteristics are changing
over time.
Inferences: Estimating Parameters and Hypothesis Testing When trying
to use sample data for making inferences about a population, it is often difficult to
choose the best procedure. Figure 15-1 includes the major methods included in this
book, along with a scheme for determining which of those methods should be used.
Figure 15-1 applies to a fixed population. If the data are from a process that may
change over time, construct a control chart (see Chapter 14) to determine whether
the process is statistically stable. Figure 15-1 applies to process data only if the process
is statistically stable. In addition to the procedures identified in Figure 15-1, there are
many other methods that might be more suitable for a particular statistical analysis.
Consult your friendly professional statistician for help with other methods.
Conclusions and Practical Implications After completing the statistical
analysis, we should state conclusions in a way that is clear to those unfamiliar with
statistics and its terminology, and we should carefully avoid making statements
not justified by the statistical analysis (such as using a correlation to conclude that
one variable is the cause of the other). Also, we should identify practical implica-
tions of the results.
 15-3 Perspectives 745

Level of Number of Claim or

Start Measurement Populations Parameter Inference

Mean Estimating
with Confidence
Interval:
Consider Variance 7-3, 7-4
• Context

•Source One Population

•Sampling method
Means: Hypothesis
Interval
9-3, 9-4 Testing
or Ratio
8-4, 8-5
(such as Two Populations
heights, Variances:
Construct graphs and weights) 9-5
find sample More Than
statistics to investigate Two Populations:
distribution and Chap. 12, 13-5 Correlation,
presence of outliers. Regression:
Chap. 10, 13-6
One Population: Estimating
13-2 with Confidence
Interval:
Independent:
What is the level of Ordinal 7-5
13-4
measurement (such as data
consisting Two Populations
of the data? Hypothesis
1-3 of ranks) Matched Pairs: Testing:
13-3 8-6
More Than
Two Populations:
13-5 Goodness-of-Fit
(one row):
Frequency 11-2
Counts for
Categories
Nominal Contingency Table
(data consisting (multiple rows,
of proportions columns):
11-3 Estimating
or frequency Proportion with
counts for Confidence
different Two Populations: Interval: 7-2
categories) 9-2
Proportions Hypothesis
Testing:
One Population 8-3, 13-2

Figure 15-1 Selecting the Appropriate Procedure

15-3 Perspectives
Key Concept No single introductory statistics course can make anyone an expert sta-
tistician. The introductory course has a limited scope, and many important topics are
not included.
Successful completion of an introductory statistics course results in benefits that
extend far beyond the attainment of credit toward a college degree. You will have im-
proved job marketability. You will be better prepared to critically analyze reports in the
media and professional journals. You will understand the basic concepts of probability
and chance. You will know that in attempting to gain insight into a set of data, it is im-
portant to consider the context of the data, the source of the data, and the sampling
methods used. You will know that, given sample data, you should investigate measures
of center (such as mean and median), measures of variation (such as range and standard
746 Chapter 15 Projects, Procedures, Perspectives

deviation), the distribution (via a frequency distribution or graph), the presence of

outliers, and whether the population is stable or is changing over time. You will know
and understand the importance of estimating population parameters (such as mean,
standard deviation, and proportion), as well as testing claims made about population
parameters.
Throughout this text we have emphasized the importance of good sampling
methods. You should recognize that a bad sample may be beyond repair by even the
most expert statisticians using the most sophisticated techniques. There are many
mail, magazine, and telephone call-in surveys that allow respondents to be “self-
selected.” The results of such surveys are generally worthless when judged according
to the criteria of sound statistical methodology. Keep this in mind when you encounter
voluntary response (self-selected) surveys, so that you don’t let them affect your be-
liefs and decisions. You should also recognize, however, that many surveys and polls
obtain very good results, even though the sample sizes might seem to be relatively
small. Although many people refuse to believe it, a nationwide survey of only 1200
voters can provide good results if the sampling is carefully planned and executed.
Throughout this text we have emphasized the interpretation of results. Comput-
ers and calculators are quite good at yielding results, but such results generally require
careful interpretation. We should recognize that a result is not automatically valid
simply because it was computer-generated. Computers don’t think, and they are quite
capable of providing results that are quite ridiculous when considered in the context
of the real world. We should always apply the most important and indispensable tool
in all of statistics: common sense!
The Educated Person Once upon a time, a person was considered to be edu-
cated if he or she could simply read. That time has long passed. Today, an educated
person is capable of critical thinking, possesses an intellectual curiosity, and can com-
municate effectively both orally and in writing. An educated person can relate to all
other people, including those from different cultures, as well as those who might not
be so educated. An educated person has statistical literacy and an ability to think sta-
tistically. Successful completion of an introductory statistics course is a great achieve-
ment of a truly educated person.