1

STAT 4013: Statistical research design and analysis for the life sciences
(2012)
Duration: 3 hours Internal examiner: Mrs. RE Krommenhoek
Total marks 180 External examiner: Dr P Becker
Full marks are 150.
All marks are approximate.

Notes: This is an open book exam.

Question 1

a) What is an estimator, give an example? (2 marks)

b) What are the basic "ingredients" of a statistical test? What are possible outcomes?
(6 marks)
c) When is the Pearson Correlation Coefficient used?
(2 marks)
d) What is the purpose of using indicator variables in a regression line equation?
(4 marks)
e) What is the purpose of the Chi Square Goodness of fit test? When is the
Kolmogorov-Smirnov used? (3 marks)
f) When is Factor Analysis used and when is Cluster Analysis used?
(2 marks)
g) What are the two main goals of time series analysis? (2 marks)
(2 marks)
h) What are the major tools used to identify the parameters for an ARIMA model?
(3 marks)
i) What are the similarities between cluster analysis and factor analysis? (2 marks)
j) What are the similarities between principal component analysis and factor analysis?
(2 marks)
k) How can a plot assist in determining whether a logistic regression should be
performed between two variables? (2 marks)
l) What is the purpose of a CHAID analysis? (2 marks)

m) How is survival analysis different from any of the other type of analyses that
generate great difficulty when trying to analyse data using traditional
statistical models such as multiple linear regression? (4 marks)
(38 marks)
Question 2

a) Suppose the 4Flu Drug Company develops a new drug, designed to prevent colds.
The company states that the drug is equally effective for men and women. To test
this claim, they choose a simple random sample of 100 women and 200 men from a
population of 100,000 volunteers.
At the end of the study, 38% of the women caught a cold; and 51% of the men
caught a cold. Based on these findings, can we reject the company's claim that the
drug is equally effective for men and women? Use a 0.05 level of significance.

2

The results are in Table 1. Discuss it in detail, define the appropriate null and
alternative hypotheses to answer the question whether the companys claim can be
rejected that the drug is equally effective for men and women? (7 marks)

b) A dentist wanted to prove that girls and boys will have the same dental
measurements at the same age. He recorded the dental measurements and
performed an analysis on them to decide if the dental measurements are the same at
the same age. Discuss the methods of analyses in Table 2 which were used and
compare the outcome of the results of the tests performed. Discuss the advantages
and disadvantages of the methods used. (9 marks)
i) Why would the dentist prefer to use parametric tests rather than non-parametric
tests, especially when the distribution is slightly skew? (2 marks)
ii) Explain the Box and Whisker plot in Figure 2.1. (2 marks)
(20 marks)
Question3

a) The KeepitCold Company manufactures refrigeration equipment as well as many
replacement parts. In the past, one of the replacement parts was produced
periodically in lots of varying sizes. When a cost improvement programme was
undertaken, company officials wished to determine the optimum lot size for
production of this part. The production of this part involved setting up the production
process (which must be done no matter what is the lot size) and machining and
assembly operations. One key input for the model was to ascertain the optimum lot
size, which was the relationship between lot size(LotSize) and labour hours
(Workhrs) required to produce the lot. To determine this relationship, data on lot
size and work hours for 25 recent production runs were utilized. The production
conditions were stable during the six-month period in which the 25 runs were made
and were expected to continue to be the same during the next three years, the
planning period for which the cost improvement program was being conducted.
Discuss the results in Table 3 in detail for LotSize and Workhrs. (3 marks)

b) A market researcher conducted a study for a portrait studio to determine what the
relationship is between sales and per capita disposable income of what people of age
16 years and younger will spend on photographic equipment. The study was done in
twenty two cities.
i) Discuss the results in Table 4.1 in detail. (14 marks)
ii) Discuss the two graphs in Figures 4.1 and 4.2 in detail. (2 marks)
iii) Discuss the residuals skewness and kurtosis in Table 4.2. (4 marks)
iv) Discuss the location test result for the residual in Table 4.2.1. (4 marks)
v) Discuss the KolmogorovSmirnov test result for normality for the residual in
Table 4.2.2. (4 marks)

(31 marks)
Question 4
a) A team of researchers is interested in determining whether two methods of hypnotic
induction, A and B, differ with respect to their effectiveness. They begin by randomly
sorting 20 volunteer subjects into two independent groups of 10 subjects each, with
the aim of administering Method A to one group and Method B to the other. But then,
before either of the induction methods is administered, each subject is pre-measured
on a standard index of "primary suggestibility," which is a variable known to be
correlated with receptivity to hypnotic induction which is the covariate. The dependent
3

variable, measured during the administration of Method A or Method B, is the
subject's score on a standard index of hypnotic induction. Effect in Table 5 is the
interaction between Methods and Scoreonsuggestibility.
i) Discuss the model, define the appropriate null and alternative hypotheses, as
well as the parameter estimates. (14 marks)
ii) Are the slopes the same for Method A and Method B? (3 marks)
iii) Are the intercepts the same for Method A and Method B? (3 marks)
iv) Is any multicollinearity present? Explain how to determine the presence of
multicollinearity. (2 marks)
v) Discuss Figures 5.1 to 5.9. (9 marks)

b) Crime rates for fifty states in America were collected in 1977 for rape, robbery,
murder, burglary, larcenry, assault and auto theft. The information is available in
Table 6.
i) Discuss in detail how the factors were obtained for Tables 6.1.1. and 6.1.2 as
well as for Table 6.2. Is it possible to differentiate between violent crimes and
non-violent crimes with factor analysis? (15 marks)
ii) Discuss Figures 6.1 and 6.2 in detail. ( 8 marks)
(54 marks)
Question 5

A senior living community developer is looking to institute new activities and
opportunities for the members of their communities in Johannesburg. They are quite
confident that those in the older age ranges will react differently than those in the
younger age ranges.

Discuss how they will be able to ensure that enough respondents will provide
meaningful comparisons. Discuss fully how you would draw a sample to determine this.
Include in your discussion the factors that you would need to consider in order to obtain
a representative sample, as well as reasons for your choice of the sampling scheme.
Also comment on your choice of method of administering the questionnaire (face-to-
face, telephone, handouts, etc), and any biases that you think could arise. (15 marks)


Question 6

Table 7 provides information on whether obtaining work experience will have an effect
on whether a person will be successful in their work environment or not.



i) Discuss the results in detail for Table 7 to answer the question whether work
experience will have an influence on success in the work environment? Define
the appropriate hypotheses. (8 marks)
ii) Interpret the odds ratio. Explain how to determine the odds ratio. (3 marks)
iii) Discuss whether the model is a good fit for this particular set of data. (5 marks)
iv) Discuss the graph in Figure 7.1 (2 marks)
v) Discuss the results in the outlier diagnostics table, Table 7.1. Determine which
observations are outliers. (4 marks)
(22 marks)

4

Table 1

Table of Cold by Gender
Gender
Total Men Women
Cold
98 62 160
0
Frequency
Col Pct 49.00 62.00
1
Frequency 102 38 140
Col Pct 51.00 38.00
Total Frequency 200 100 300
Statistics for Table of Cold by Gender
Column 1 Risk Estimates
Risk ASE
(Asymptotic) 95%
Confidence Limits
(Exact) 95%
Confidence Limits
Row 1 0.6125 0.0385 0.5370 0.6880 0.5324 0.6884
Row 2 0.7286 0.0376 0.6549 0.8022 0.6470 0.8002
Total 0.6667 0.0272 0.6133 0.7200 0.6102 0.7198
Difference -0.1161 0.0538 -0.2215 -0.0106
Difference is (Row 1 - Row 2)
Column 2 Risk Estimates
Risk ASE
(Asymptotic) 95%
Confidence Limits
(Exact) 95%
Confidence Limits
Row 1 0.3875 0.0385 0.3120 0.4630 0.3116 0.4676
Row 2 0.2714 0.0376 0.1978 0.3451 0.1998 0.3530
Total 0.3333 0.0272 0.2800 0.3867 0.2802 0.3898
Difference 0.1161 0.0538 0.0106 0.2215
Difference is (Row 1 - Row 2)
Sample Size = 300



















5


Table 2
The NPAR1WAY Procedure
Wilcoxon Scores (Rank Sums) for Variable mm
Classified by Variable gender
gender N
Sum of
Scores
Expected
Under H0
Std Dev
Under H0
Mean
Score
girls 11 91.50 154.0 20.202959 8.318182
boys 16 286.50 224.0 20.202959 17.906250
Average scores were used for ties.
Wilcoxon Two-Sample Test
Statistic (S) 91.5000

Normal Approximation
Z -3.0689
One-Sided Pr < Z 0.0011
Two-Sided Pr > |Z| 0.0021

t Approximation
One-Sided Pr < Z 0.0025
Two-Sided Pr > |Z| 0.0050

Exact Test
One-Sided Pr <= S 6.068E-04
Two-Sided Pr >= |S - Mean| 0.0012
Z includes a continuity correction of 0.5.
Kruskal-Wallis Test
Chi-Square 9.5704
DF 1
Pr > Chi-Square 0.0020

t Test
The TTEST Procedure
Variable: DentalMeasurements
Gender N Mean Std Dev Std Err Minimum Maximum
1 11 24.0909 2.4374 0.7349 19.5000 28.0000
2 16 27.4688 2.0854 0.5214 25.0000 31.5000
Diff (1-2) -3.3778 2.2329 0.8746
Gender Method Mean 95% CL Mean Std Dev 95% CL Std Dev
1 24.0909 22.4534 25.7284 2.4374 1.7031 4.2775
2 27.4688 26.3575 28.5800 2.0854 1.5405 3.2276
Diff (1-2) Pooled -3.3778 -5.1790 -1.5766 2.2329 1.7511 3.0823
Diff (1-2) Satterthwaite -3.3778 -5.2616 -1.4941
6

Method Variances DF t Value Pr > |t|
Pooled Equal 25 -3.86 0.0007
Satterthwaite Unequal 19.334 -3.75 0.0013
Equality of Variances
Method Num DF Den DF F Value Pr > F
Folded F 10 15 1.37 0.5668

Figure 2.1






















7


Table 3

The CORR Procedure
2 Variables: WorkHrs LotSize
Simple Statistics
Variable N Mean Std Dev Sum Minimum Maximum
WorkHrs 25 312.28000 113.13764 7807 113.00000 546.00000
LotSize 25 70.00000 28.72281 1750 20.00000 120.00000
Pearson Correlation Coefficients, N = 25
Prob > |r| under H0: Rho=0
WorkHrs LotSize
WorkHrs
1.00000

0.90638
<.0001
LotSize
0.90638
<.0001
1.00000


Pearson Correlation Statistics (Fisher's z Transformation)
Variable
With
Variable N
Sample
Correlation
Fisher's
z
Bias
Adjustment
Correlation
Estimate 95% Confidence Limits
p Value
for
H0:Rho=0
WorkHrs LotSize 25 0.90638 1.50689 0.01888 0.90296 0.789513 0.956737 <.0001

























8


Table 4

Table 4.1


The REG Procedure

Model: MODEL1

Dependent Variable: sales
Number of Observations Read 22
Number of Observations Used 21
Number of Observations with Missing Values 1
Analysis of Variance
Source DF
Sum of
Squares
Mean
Square F Value Pr > F
Model 2 24015 12008 99.10 <.0001
Error 18 2180.92741 121.16263
Corrected Total 20 26196
Root MSE 11.00739 R-Square 0.9167
Dependent Mean 181.90476 Adj R-Sq 0.9075
Coeff Var 6.05118
Parameter Estimates
Variable DF
Parameter
Estimate
Standard
Error t Value Pr > |t|
Intercept 1 -68.85707 60.01695 -1.15 0.2663
people16 1 1.45456 0.21178 6.87 <.0001
income 1 9.36550 4.06396 2.30 0.0333




















9

The REG Procedure
Model: MODEL1
Dependent Variable: sales
Obs
Dependent
Variable
Predicted
Value
Std Error
Mean Predict Residual
Std
Error
Residual
Student
Residual
Cook's
D




1 174.4000 187.1841 3.8409 -12.7841 10.316 -1.239 0.071
2 164.4000 154.2294 3.5558 10.1706 10.417 0.976 0.037
3 244.2000 234.3963 4.5882 9.8037 10.006 0.980 0.067
4 154.6000 153.3285 3.2331 1.2715 10.522 0.121 0.000
5 181.6000 161.3849 4.4300 20.2151 10.077 2.006 0.259
6 207.5000 197.7414 4.3786 9.7586 10.099 0.966 0.059
7 152.8000 152.0551 4.1696 0.7449 10.187 0.0731 0.000
8 163.2000 167.8666 3.3310 -4.6666 10.491 -0.445 0.007
9 145.4000 157.7382 2.9628 -12.3382 10.601 -1.164 0.035
10 137.2000 136.8460 4.0074 0.3540 10.252 0.0345 0.000
11 241.9000 230.3874 4.2012 11.5126 10.174 1.132 0.073
12 191.1000 197.1849 3.4109 -6.0849 10.466 -0.581 0.012
13 232.0000 222.6857 5.3808 9.3143 9.603 0.970 0.098
14 145.3000 141.5184 4.1735 3.7816 10.186 0.371 0.008
15 161.1000 174.2132 5.0377 -13.1132 9.787 -1.340 0.159
16 209.7000 228.1239 4.1214 -18.4239 10.207 -1.805 0.177
17 146.4000 145.7470 3.7331 0.6530 10.355 0.0631 0.000
18 144.0000 159.0013 3.2529 -15.0013 10.516 -1.427 0.065
19 232.6000 230.9870 4.4176 1.6130 10.082 0.160 0.002
20 224.1000 230.3161 5.8120 -6.2161 9.348 -0.665 0.057
21 166.5000 157.0644 4.0792 9.4356 10.224 0.923 0.045
22 . 191.1039 2.7668 . . . .
Sum of Residuals 0
Sum of Squared Residuals 2180.92741
Predicted Residual SS (PRESS) 3002.92331















10

Figure 4.1




















11

Figure 4.2



Table 4.2
The UNIVARIATE Procedure

Variable: RES (Residual)
Moments
N 21 Sum Weights 21
Mean 0 Sum Observations 0
Std Deviation 10.442527 Variance 109.046371
Skewness -0.0970495 Kurtosis -0.7942686
Uncorrected SS 2180.92741 Corrected SS 2180.92741
Coeff Variation . Std Error Mean 2.27874622
Basic Statistical Measures
Location Variability
Mean 0.000000 Std Deviation 10.44253
Median 0.744918 Variance 109.04637
Mode . Range 38.63896
Interquartile Range 15.65166
12

Table 4.2.1
Tests for Location: Mu0=0
Test Statistic p Value
Student's t t 0 Pr > |t| 1.0000
Sign M 2.5 Pr >= |M| 0.3833
Signed Rank S 1.5 Pr >= |S| 0.9599
Table 4.2.2
Tests for Normality
Test Statistic p Value
Shapiro-Wilk W 0.954073 Pr < W 0.4056
Kolmogorov-Smirnov D 0.147126 Pr > D >0.1500
Cramer-von Mises W-Sq 0.066901 Pr > W-Sq >0.2500
Anderson-Darling A-Sq 0.432299 Pr > A-Sq >0.2500





































13

Table 5
Linear Regression Results

Dependent Variable: Hypnotic induction
Number of Observations Read 20
Number of Observations Used 20
Analysis of Variance
Source DF
Sum of
Squares
Mean
Square F Value Pr > F
Model 3 556.81966 185.60655 26.58 <.0001
Error 16 111.73034 6.98315
Corrected Total 19 668.55000
Root MSE 2.64256 R-Square 0.8329
Dependent Mean 28.65000 Adj R-Sq 0.8015
Coeff Var 9.22361
Parameter Estimates
Variable DF
Parameter
Estimate
Standard
Error t Value Pr > |t|
Intercept 1 18.51156 1.80617 10.25 <.0001
Methods 1 -3.40287 2.97979 -1.14 0.2703
Effects(Interaction of Methods with Scoreonsuggestibility) 1 -0.12655 0.17700 -0.71 0.4849
Scoreonsuggestibility 1 0.84829 0.12708 6.68 <.0001
Collinearity Diagnostics (intercept adjusted)
Number Eigenvalue
Condition
Index
Proportion of Variation
Methods Effects Scoreonsuggestibility
1 2.28989 1.00000 0.02346 0.01949 0.04401
2 0.64804 1.87977 0.07330 0.00285 0.43754
3 0.06207 6.07388 0.90324 0.97766 0.51846


















14


Figure 5.1


















15

Figure 5.2



















16

Figure 5.3



















17

Figure 5.4



















18

Figure 5.5



















19

Figure 5.6



















20

Figure 5.7



















21

Figure 5.8



















22

Figure 5.9
























23


Table 6
Factor Analysis Results
The FACTOR Procedure
Input Data Type Raw Data
Number of Records Read 50
Number of Records Used 50
N for Significance Tests 50

Table 6.1

Table 6.1.1

Factor Analysis Results
The FACTOR Procedure
Initial Factor Method: Principal Components
Prior Communality Estimates: ONE
Eigenvalues of the Correlation Matrix: Total
= 7 Average = 1
Eigenvalue Difference Proportion Cumulative
1 4.11495951 2.87623768 0.5879 0.5879
2 1.23872183 0.51290521 0.1770 0.7648
3 0.72581663 0.40938458 0.1037 0.8685
4 0.31643205 0.05845759 0.0452 0.9137
5 0.25797446 0.03593499 0.0369 0.9506
6 0.22203947 0.09798342 0.0317 0.9823
7 0.12405606 0.0177 1.0000









24


Figure 6.1


Table 6.1.2
Factor Pattern
Factor1 Factor2
Rape 0.87584 -0.18858
Robbery 0.80508 0.04702
Murder 0.60913 -0.70026
Burglary 0.89287 0.22631
Larcenry 0.72492 0.44777
Assault 0.80462 -0.38234
Auto 0.59878 0.55918
Variance Explained by Each
Factor
Factor1 Factor2
4.1149595 1.2387218
Final Communality Estimates: Total = 5.353681
Rape Robbery Murder Burglary Larcenry Assault Auto
0.80265644 0.65035880 0.86139693 0.84844347 0.72600469 0.79360069 0.67122032





25




Table 6.2

Factor Analysis Results
The FACTOR Procedure
Rotation Method: Varimax
Orthogonal Transformation Matrix
1 2
1 0.72162 0.69228
2 0.69228 -0.72162
Rotated Factor Pattern
Factor1 Factor2
Rape 0.50148 0.74241
Robbery 0.61351 0.52341
Murder -0.04522 0.92701
Burglary 0.80099 0.45481
Larcenry 0.83310 0.17873
Assault 0.31595 0.83293
Auto 0.81921 0.01100
Variance Explained by Each
Factor
Factor1 Factor2
2.7364991 2.6171822
Final Communality Estimates: Total = 5.353681
Rape Robbery Murder Burglary Larcenry Assault Auto
0.80265644 0.65035880 0.86139693 0.84844347 0.72600469 0.79360069 0.67122032

















26



Factor Analysis Results
The FACTOR Procedure
Rotation Method: Varimax
Figure 6.2


















27



Table 7


The LOGISTIC Procedure
Model Information
Data Set WORK.PROGTASK
Response Variable success
Number of Response Levels 2
Model binary logit
Optimization Technique Fisher's scoring
Number of Observations Read 25
Number of Observations Used 25
Response Profile
Ordered
Value success
Total
Frequency
1 1 11
2 0 14
Probability modeled is success=1.
Model Fit Statistics
Criterion
Intercept
Only
Intercept
and
Covariates
AIC 36.296 29.425
SC 37.515 31.862
-2 Log L 34.296 25.425
Testing Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood Ratio 8.8719 1 0.0029
Score 7.9742 1 0.0047
Wald 6.1760 1 0.0129
Analysis of Maximum Likelihood Estimates
Parameter DF Estimate
Standard
Error
Wald
Chi-Square Pr > ChiSq
Intercept 1 -3.0597 1.2594 5.9029 0.0151
exper 1 0.1615 0.0650 6.1760 0.0129
Odds Ratio Estimates
Effect Point Estimate
95% Wald
Confidence Limits
exper 1.175 1.035 1.335
Association of Predicted Probabilities and
Observed Responses
Percent Concordant 82.5 Somers' D 0.662
Percent Discordant 16.2 Gamma 0.671
Percent Tied 1.3 Tau-a 0.340
28

Pairs 154 c 0.831
Partition for the Hosmer and Lemeshow Test
Group Total
success = 1 success = 0
Observed Expected Observed Expected
1 3 0 0.26 3 2.74
2 3 1 0.37 2 2.63
3 3 0 0.63 3 2.37
4 3 1 0.86 2 2.14
5 3 1 1.43 2 1.57
6 3 3 1.78 0 1.22
7 3 2 2.23 1 0.77
8 4 3 3.44 1 0.56
Hosmer and Lemeshow Goodness-of-Fit
Test
Chi-Square DF Pr > ChiSq
5.1453 6 0.5253









]
























29


Figure 7.1







Table 7.1: Outlier diagnostics table
Case
Number
Covariates
Pearson
Residual
Deviance
Residual
Hat Matrix
Diagonal
Intercept
DfBeta
Value
(1
unit
=
0.09)
-8
-4 0
2 4 6
8

exper DfBeta
Value

(1
unit
=
0.08)
-8
-4 0
2 4 6
8


exper

Value

(1
unit
=
0.3)
-8
-4
0 2
4 6
8

Value

(1
unit
=
0.25)
-8
-4 0
2 4 6
8

Value

(1 unit
= 7.E-
03)
0 2 4
6 8 12
16


1 14.0000 -0.6707 |
* |
-0.8619 |
* |
0.0647 |
*
-0.1397 |
* |
0.0832 |
|*
30

Case
Number
Covariates
Pearson
Residual
Deviance
Residual
Hat Matrix
Diagonal
Intercept
DfBeta
Value
(1
unit
=
0.09)
-8
-4 0
2 4 6
8

exper DfBeta
Value

(1
unit
=
0.08)
-8
-4 0
2 4 6
8


exper

Value

(1
unit
=
0.3)
-8
-4
0 2
4 6
8

Value

(1
unit
=
0.25)
-8
-4 0
2 4 6
8

Value

(1 unit
= 7.E-
03)
0 2 4
6 8 12
16


|

|

|

|

|


2
29.0000 -2.2517 | *
|
|

-1.8992 |*
|
|

0.1051 |
*|

0.4452 |
|
*
|

-0.6758 |*
|
|


3
6.0000 -0.3516 |
*|
|

-0.4828 |
* |
|

0.0816 |
* |

-0.1078 |
*|
|

0.0923 |
|*
|


4
25.0000 0.6134 |
| *
|

0.7992 |
| *
|

0.0904 |
* |

-0.0714 |
*|
|

0.1395 |
| *
|


5
18.0000 1.0795 |
|
*
|

1.2430 |
|
* |

0.0595 |
*
|

0.1039 |
|*
|

0.00529 |
*
|


6
4.0000 -0.2991 |
*|
|

-0.4140 |
* |
|

0.0792 |
* |

-0.0910 |
*|
|

0.0803 |
|*
|


7
18.0000 -0.9264 |
* |
|

-1.1132 | *
|
|

0.0595 |
*
|

-0.0892 |
*|
|

-0.00454 |
*
|


8
12.0000 -0.5707 |
* |
|

-0.7509 |
* |
|

0.0712 |
* |

-0.1431 |
* |
|

0.1007 |
|*
|


9
22.0000 0.7815 |
| *
|
0.9765 |
| *
|
0.0734 |
* |

-0.0237 |
*
|
0.1102 |
|*
|
31

Case
Number
Covariates
Pearson
Residual
Deviance
Residual
Hat Matrix
Diagonal
Intercept
DfBeta
Value
(1
unit
=
0.09)
-8
-4 0
2 4 6
8

exper DfBeta
Value

(1
unit
=
0.08)
-8
-4 0
2 4 6
8


exper

Value

(1
unit
=
0.3)
-8
-4
0 2
4 6
8

Value

(1
unit
=
0.25)
-8
-4 0
2 4 6
8

Value

(1 unit
= 7.E-
03)
0 2 4
6 8 12
16




10
6.0000 -0.3516 |
*|
|

-0.4828 |
* |
|

0.0816 |
* |

-0.1078 |
*|
|

0.0923 |
|*
|


11
30.0000 0.4097 |
|*
|

0.5570 |
| *
|

0.1061 |
*|

-0.0863 |
*|
|

0.1268 |
| *
|


12
11.0000 -0.5264 |
* |
|

-0.6994 |
* |
|

0.0744 |
* |

-0.1408 |
* |
|

0.1045 |
|*
|


13
30.0000 0.4097 |
|*
|

0.5570 |
| *
|

0.1061 |
*|

-0.0863 |
*|
|

0.1268 |
| *
|


14
5.0000 -0.3243 |
*|
|

-0.4472 |
* |
|

0.0807 |
* |

-0.0994 |
*|
|

0.0865 |
|*
|


15
20.0000 0.9185 |
| *
|

1.1061 |
|
* |

0.0642 |
*
|

0.0303 |
*
|

0.0681 |
|*
|


16
13.0000 -0.6187 |
* |
|

-0.8051 |
* |
|

0.0679 |
*
|

-0.1428 |
* |
|

0.0937 |
|*
|


17
9.0000 -0.4479 |
*|
|

-0.6047 |
* |
|

0.0794 |
* |

-0.1308 |
*|
|

0.1045 |
|*
|


18
32.0000 0.3486 |
|*
0.4789 |
| *
0.1047 |
*|
-0.0795 |
*|
0.1112 |
|*
32

Case
Number
Covariates
Pearson
Residual
Deviance
Residual
Hat Matrix
Diagonal
Intercept
DfBeta
Value
(1
unit
=
0.09)
-8
-4 0
2 4 6
8

exper DfBeta
Value

(1
unit
=
0.08)
-8
-4 0
2 4 6
8


exper

Value

(1
unit
=
0.3)
-8
-4
0 2
4 6
8

Value

(1
unit
=
0.25)
-8
-4 0
2 4 6
8

Value

(1 unit
= 7.E-
03)
0 2 4
6 8 12
16


|

|

|

|


19
24.0000 -1.5038 |
*
|
|

-1.5376 | *
|
|

0.0848 |
* |

0.1346 |
|*
|

-0.3024 |
* |
|


20
13.0000 1.6164 |
|
*
|

1.6028 |
|
* |

0.0679 |
*
|

0.3732 |
|
*
|

-0.2449 |
* |
|


21
19.0000 -1.0043 |
* |
|

-1.1810 | *
|
|

0.0612 |
*
|

-0.0651 |
*|
|

-0.0397 |
*
|


22
4.0000 -0.2991 |
*|
|

-0.4140 |
* |
|

0.0792 |
* |

-0.0910 |
*|
|

0.0803 |
|*
|


23
28.0000 0.4814 |
| *
|

0.6457 |
| *
|

0.1029 |
*|

-0.0875 |
*|
|

0.1382 |
| *
|


24
22.0000 0.7815 |
| *
|

0.9765 |
| *
|

0.0734 |
* |

-0.0237 |
*
|

0.1102 |
|*
|


25
8.0000 2.4203 |
|
*|

1.9624 |
|
*|

0.0809 |
* |

0.7253 |
|
*|

-0.5953 | *
|
|