I88 Chapter l7 Statistics for Quality: Control and Capability
Example 17.6 Measuring Clips. The dimension of the opening of a clip has
specifications 15 i 0.5 millimeters. The production of the clip is monitored by a? and 3
charts based on samples of five

CHAPTER
11
'I'1=FI=-:|FIIH:5:'-III
H=E.'-I255315| .'|'=.1|I|55155|3 :
Multiple
Regression
11.1 Inference for Multiple Regression
11.2 A Case Study
Introduction
In this chapter, we demonstrate how to use a program for Tl-83 and -84 calculators

I 16 Chapter IO Inference for Regression
t=?. 4?2358
dF=14
F-=2. 9694F'5'921 E 'E-
E_ LE'u'EL=. 95
Last, the intercept is not significantly different from 0, since 0
is included in the confidence interval for the value when x: 0. M_
This makes sense;

I38 Chapter [3 Two Way Analysis of Variance
13.1 Plotting Means
Example 13.1 Time Spent Eating. The table below gives the mean length of time (in
minutes) that various groups of people spent eating lunch in various settings. Plot the
group means for this

More Details about Simple Linear Regression l2l
The REGINF Program for Linear Regression Inference
:Input "X LIST=",LX
:Input Y LlST=",LY
:FnOff
:LinRegTTest LX,LY,0,Y1 :
16X2 "-VZSSZiMi -N
:Cerome
:Output(1,2,Y=a+bX"
:Output(2,2,"a="
:Output(2,4,a
:Outpu

I28 Chapter I I Multiple Regression
Solution. First, we must edit the last column of matrix [A] so that it contains the SATM
scores. To enter these easily, use the down arrow after each entry rather than pressing
-. Then rerun the MU LREG program to obtai

I32 Chapter [2 One Way Analysis of Variance
12.1 Inference for One-Way Analysis of Variance
We begin with an exercise that demonstrates built-in analysis EDIT CHLE
of variance capabilities using the ANOVA( command from the ETEEEEEEIH
STAT TESTS menu. HEEU

I 18 Chapter IO Inference for Regression
Solution. One particular home is not an average. In this case,
we need a prediction interval for a new observation. We
found all the needed pieces in Example 10.5. All we need to
do is add another l under the squar

120 Chapter IO Inference for Regression
t = 1.266220727. If we square this value, then we obtain 1.603314929, which is the
actual value of the displayed (rounded) F -statistic from the ANOVA test.
Sample Correlation and the tTest
One may be required to pe

Inference for Two-Way ANOVA I43
Unfortunately, the Tl89 does not create the cell means for you to create the means plot.
You can calculate these fairly easily, though. Enter four 1s in list 4, then four 23, then
four 3 5. These will correspond to the type

A Case Study 129
(a) Use a simple linear regression to predict assets using the number of accounts. Give
the regression equation and the results of the significance test for the regression
coefficient.
(b) Do the same using market share to predict assets.

I 14 Chapter IO Inference for Regression
You are now asked for the confidence level. Here we want 90%, so enter .9. You are
now asked what x value we are interested in. Since we are interested in the intercept,
enter 0. Press to continue. We first find a

Using Control Charts I85
We notice that there were four samples out of control on the 27 chart: the third, sixth,
tenth, and seventeenth. There were two samples (the first and sixth) out of control on the
s chart.
(b) Because the upper control limit for s

I86 Chapter l7 Statistics for Quality: Control and Capability
Solution. Simply make a time plot of the measurements together with graphs of the lines
y = ,u and y = ,ui2a. To do so, enter the weeks into list L1 and the weights into list
L2, then adjust th

190 Chapter l7 Statistics for Quality: Control and Capability
Example 17.8 School Absenteeism. Here are data on the total number of absentees
among eighth-graders with three or more unexcused absences at an urban school district.
Because the total number

I80 Chapter l7 Statistics for Quality: Control and Capability
17.1 Statistical Process Control
In this section, we provide a program that computes the upper and lower control limits
and graphs the control charts for f and s.
The CONTRL Pro ram
PROGRAMCONT

CHAPTER
17
Statistics for
Quality: Control
and Capability
17.1 Statistical Process Control
17.2 Using Control Charts
17.3 Process Capability Indexes
17.4 Control Charts for Sam 1e Pro ortions
Introduction
In this chapter, we provide several programs f

Control Charts for Sample Proportions I89
The CONTRLP Pro ram
PROGRAMCONTRLP
:Menu("CONTRLP","STATS",1,LIST",2,"
QUIT",3)
:Lbl 1
:Disp "TOTAL SUCCESSES"
:lnput T
:Disp "NO. OF STAGES"
:Input M
:Disp "NO. PER STAGE
:Input N
:T/(M*N)-P
:Goto 4
:Lb12
:l-Var

192 Chapter [8 Time Series Forecasting
18.1 Trends and Seasons
The most basic component of a time series is assessing whether or not there is a trend
(systematic rise or fall) and whether there is some aspect that repeats regularly (a cycle,
or seasonal c

Process Capability Indexes 187
Example 17.5 Hospital Losses. Below are data on a hospitals losses for 120 DRG 209
(major joint replacement) patients collected as 15 monthly samples of eight patients each.
The hospital has determined that suitable specific

The Logistic Regression Model I73
16.1 The Logistic Regression Model
First, we provide a supplementary program that computes appropriate mathematical odds
for a given probability p of an event A. If p S 0.50, then the odds against A are given as
the ratio

Statistical Process Control [8|
Example 17.1 Milling Hydraulic Systems. The width ofa slot cut by a milling machine
is important for the proper functioning of a hydraulic system for large tractors. The
manufacturer checks the control of the milling proces

CHAPTER
Y
O 0
Lo g1 st1c
0
Re gre s 5 ion
16.1 The Logistic Regression Model
16.2 Inference for Lo istic Re ression
Introduction
In this chapter, we give a brief discussion of two types of logistic regression fits. The
first type is a linear t for the

The Wilcoxon Signed Rank Test 165
Unlogged 22 18 22 20 15 21 13 l3 l9 l3 19 15
Logged 17 4 18 14 18 15 15 10 12
Does logging signicantly reduce the mean (median) number of species in a plot after
eight years? State the hypotheses, do a Wilcoxon rank sum

CHAPTER
W
Time Series
Forecasting
18.1
18.2
Trends and Seasons
Time Series Models
Introduction
A time series is a sequence of observations on a single variable at equally spaced
intervals.
In this chapter, we examine some basic ideas in time ser

152 Chapter l4 Bootstrap Methods and Permutation Tests
With the sample in list L1 , we can execute the BOOT program to bootstrap the sample
mean. In this case, we enter 0 for CONF. INTERVAL? because we are only interested
here in the bootstrap distribut

I42 Chapter [3 Two Way Analysis of Variance
To create the means plot, use Matr>Li st from the Matrix Math menu to create lists
to plot. You need to specify as many lists as there are columns in [B]. Next, input a
column containing values 1, 2, and 3 in L4

First Steps in Using the Bootstrap [5|
in mean reading scores for the two teaching methods of (8.177, 11.732). Since the
interval does not contain 0, we have evidence that the directed reading activities help
students reading ability.
PPEIP'IEUUTF'FIIR EI

Inference for One-Way Analysis of Variance I33
The second assumption is that the data come from Normal populations. We could do a
Normal plot for each type of ower, but in this case side-by-side boxplots are a good
idea. Were looking for indications of sk