The SAS System

Directory
Libref LIBRARY
Engine V9
Physical Name C:\MyGithub\N736Fall2017_HELPdataset
Filename C:\MyGithub\N736Fall2017_HELPdataset

# Name Member Type File Size Last Modified
1 FORMATS CATALOG 33792 11/05/2017 19:28:41
2 HELP DATA 175104 09/17/2017 06:30:26
3 HELPMKH DATA 361472 11/05/2017 13:53:38


The SAS System

The UNIVARIATE Procedure
Variable: d1 (How many times hospitalized for medical problems (lifetime))

Moments
N 453 Sum Weights 453
Mean 3.05960265 Sum Observations 1386
Std Deviation 6.18758941 Variance 38.2862627
Skewness 9.78164035 Kurtosis 137.817594
Uncorrected SS 21546 Corrected SS 17305.3907
Coeff Variation 202.235065 Std Error Mean 0.29071831

Basic Statistical Measures
Location Variability
Mean 3.059603 Std Deviation 6.18759
Median 2.000000 Variance 38.28626
Mode 1.000000 Range 100.00000
    Interquartile Range 2.00000

Tests for Location: Mu0=0
Test Statistic p Value
Student's t t 10.52429 Pr > |t| <.0001
Sign M 180.5 Pr >= |M| <.0001
Signed Rank S 32670.5 Pr >= |S| <.0001

Quantiles (Definition 5)
Level Quantile
100% Max 100
99% 20
95% 10
90% 7
75% Q3 3
50% Median 2
25% Q1 1
10% 0
5% 0
1% 0
0% Min 0

Extreme Observations
Lowest Highest
Value Obs Value Obs
0 451 20 275
0 450 22 2
0 448 36 286
0 442 40 175
0 437 100 429

Plots for d1


The SAS System

The UNIVARIATE Procedure

Histogram for d1


The SAS System

The UNIVARIATE Procedure
Fitted Normal Distribution for d1 (How many times hospitalized for medical problems (lifetime))

Parameters for Normal Distribution
Parameter Symbol Estimate
Mean Mu 3.059603
Std Dev Sigma 6.187589

Goodness-of-Fit Tests for Normal Distribution
Test Statistic p Value
Kolmogorov-Smirnov D 0.3104857 Pr > D <0.010
Cramer-von Mises W-Sq 12.7755204 Pr > W-Sq <0.005
Anderson-Darling A-Sq 66.0613233 Pr > A-Sq <0.005

Quantiles for Normal Distribution
Percent Quantile
Observed Estimated
1.0 0.0000 -11.33488
5.0 0.0000 -7.11808
10.0 0.0000 -4.87011
25.0 1.0000 -1.11386
50.0 2.0000 3.05960
75.0 3.0000 7.23307
90.0 7.0000 10.98932
95.0 10.0000 13.23728
99.0 20.0000 17.45409


The SAS System

The GENMOD Procedure

Model Information
Data Set WORK.HELPMKH  
Distribution Poisson  
Link Function Log  
Dependent Variable d1 How many times hospitalized for medical problems (lifetime)

Number of Observations Read 453
Number of Observations Used 453

Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 451 1899.7021 4.2122
Scaled Deviance 451 1899.7021 4.2122
Pearson Chi-Square 451 3178.6758 7.0481
Scaled Pearson X2 451 3178.6758 7.0481
Log Likelihood   345.0367  
Full Log Likelihood   -1457.0324  
AIC (smaller is better)   2918.0647  
AICC (smaller is better)   2918.0914  
BIC (smaller is better)   2926.2965  

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% Confidence Limits Wald Chi-Square Pr > ChiSq
Intercept 1 3.2060 0.1059 2.9984 3.4135 916.66 <.0001
pcs 1 -0.0461 0.0024 -0.0508 -0.0414 365.06 <.0001
Scale 0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held fixed.


LR Statistics For Type 3 Analysis
Source DF Chi-Square Pr > ChiSq
pcs 1 362.19 <.0001


The SGPlot Procedure


The SGPlot Procedure


The SAS System

The GENMOD Procedure

Model Information
Data Set WORK.HELPMKH  
Distribution Negative Binomial  
Link Function Log  
Dependent Variable d1 How many times hospitalized for medical problems (lifetime)

Number of Observations Read 453
Number of Observations Used 453

Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 451 475.6527 1.0547
Scaled Deviance 451 475.6527 1.0547
Pearson Chi-Square 451 730.0271 1.6187
Scaled Pearson X2 451 730.0271 1.6187
Log Likelihood   817.4048  
Full Log Likelihood   -984.6643  
AIC (smaller is better)   1975.3286  
AICC (smaller is better)   1975.3820  
BIC (smaller is better)   1987.6762  

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% Confidence Limits Wald Chi-Square Pr > ChiSq
Intercept 1 3.1266 0.2324 2.6710 3.5821 180.96 <.0001
pcs 1 -0.0444 0.0049 -0.0539 -0.0348 83.30 <.0001
Dispersion 1 0.9105 0.0855 0.7575 1.0944    

Note: The negative binomial dispersion parameter was estimated by maximum likelihood.


LR Statistics For Type 3 Analysis
Source DF Chi-Square Pr > ChiSq
pcs 1 80.70 <.0001


The SGPlot Procedure


The SGPlot Procedure


The SGPlot Procedure


The GENMOD Procedure

Model Information
Data Set WORK.HELPMKH  
Distribution Poisson  
Link Function Log  
Dependent Variable d1 How many times hospitalized for medical problems (lifetime)

Number of Observations Read 453
Number of Observations Used 453

Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 452 2261.8893 5.0042
Scaled Deviance 452 2261.8893 5.0042
Pearson Chi-Square 452 5656.0908 12.5135
Scaled Pearson X2 452 5656.0908 12.5135
Log Likelihood   163.9431  
Full Log Likelihood   -1638.1260  
AIC (smaller is better)   3278.2519  
AICC (smaller is better)   3278.2608  
BIC (smaller is better)   3282.3678  

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% Confidence Limits Wald Chi-Square Pr > ChiSq
Intercept 1 1.1183 0.0269 1.0656 1.1709 1733.28 <.0001
Scale 0 1.0000 0.0000 1.0000 1.0000    

Note: The scale parameter was held fixed.



The GENMOD Procedure

Model Information
Data Set WORK.HELPMKH  
Distribution Negative Binomial  
Link Function Log  
Dependent Variable d1 How many times hospitalized for medical problems (lifetime)

Number of Observations Read 453
Number of Observations Used 453

Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 452 481.8317 1.0660
Scaled Deviance 452 481.8317 1.0660
Pearson Chi-Square 452 1235.3462 2.7331
Scaled Pearson X2 452 1235.3462 2.7331
Log Likelihood   777.0562  
Full Log Likelihood   -1025.0128  
AIC (smaller is better)   2054.0257  
AICC (smaller is better)   2054.0523  
BIC (smaller is better)   2062.2574  

Algorithm converged.

Analysis Of Maximum Likelihood Parameter Estimates
Parameter DF Estimate Standard Error Wald 95% Confidence Limits Wald Chi-Square Pr > ChiSq
Intercept 1 1.1183 0.0575 1.0056 1.2309 378.57 <.0001
Dispersion 1 1.1696 0.1010 0.9876 1.3852    

Note: The negative binomial dispersion parameter was estimated by maximum likelihood.



The SGPlot Procedure