|
Наш адрес: |
Выбрав любое изображение, кликните по нему мышкой, и Вы узнаете о статистике ...
|
|
Прочитайте ПОПУЛЯРНОЕ МНЕНИЕ МЕДИКА ИЗ ШАНХАЯ (КИТАЙ) О БИОСТАТИСТИКЕ...
Пример логистической регрессии
В. Леонов
Уважаемый автор базы данных! Перед изучением этих результатов по проверке нормальности распределения количественных признаков в сравниваемых подгруппах, Вам необходимо вначале внимательно ознакомиться с подробным описанием таких результатов в высланных Вам файлах «логит-регрессия_1.docx», « логит-регрессия_2.doc», «логит-регрессия_3.docx», « Логит-регрессия_4.rtf» с описаниями этих результатов. Детали этих результатов также объясню Вам в нашей беседе по Скайпу. Напоминаю, что перед нашей беседой сделайте в Скайпе запись сеанса в видео-файле. И тогда при необходимости Вы сможете ещё несколько раз прослушать нашу беседу через этот видео. Примеры описания результатов по логистической регрессии в 27 журнальных статьях, диссертациях и примерах этих уравнений в EXCEL-файлах, с вычислением вероятностей, которые можете прочитать в архиве с адресом, указанные Вам в письме с этим результатом. Также примеры использования и описания результатов логистической регрессии представлены в диссертациях, представленных на сайте БИОМЕТРИКА (http://www.biometrica-tomsk.ru/ ). Также в конце файла приводятся графики с ROC-кривыми, отражающие продуктивность этого уравнения.
N.B. Для зависимого признака CURE3Q были установлены 17
уравнений. В этих файлах с 17 уравнениями общий объём содержит 2519 Word-страниц. А для
демонстрации в этом файле приведён такой результат только от 45 страниц.
Логит-регрессия для двух групп CURE3Q в БД BAZA3 по п.96
Программы 3
НЦ BIOSTATISTICA. Skype: leo_1911 E-mail: leo.biostat@gmail.com
Web-Site BIOMETRICA http:\\www.biometrica-tomsk.ru\ Tel: 8-989-830-83-10
------------------------------------- CURE3Q=1 --------------------------------------
The MEANS Procedure
Variable N
---------------
CURE3Q 127
CURE1 127
CURE2Q 127
CURE4 127
CURE5 127
CURE6 127
CURE7Q 127
CURE8 127
CURE9Q 127
CURE10 127
CURE11 127
CURE12 127
CURE13 127
CURE14 127
CURE15 127
CURE16 127
CURE17 127
CURE18 127
CURE19 127
CURE20 127
CURE21 127
CURE22 127
CURE23 127
CURE24 127
CURE25 127
CURE26 127
CURE27 127
CURE28 127
CURE29Q 127
CURE30 127
CURE31 127
CURE32 127
CURE33 127
CURE34 127
CURE35 127
CURE36 127
CURE37 127
CURE38 127
CURE39 127
CURE40 127
CURE41 127
CURE42 127
CURE43Q 127
CURE44 127
CURE45 127
CURE46 127
CURE47 127
CURE48 127
CURE49 127
CURE50 127
CURE51 127
CURE52 127
CURE53 127
CURE54 127
CURE55 127
CURE56 127
CURE57 127
CURE58 127
CURE59 127
CURE60 127
CURE61 127
CURE62 127
CURE63 127
CURE64 117
CURE65 117
CURE66 117
CURE67 117
CURE68 117
CURE69 117
CURE70 117
CURE71 117
CURE72 117
CURE73 117
CURE74 117
CURE75 127
CURE76 127
CURE77 127
CURE78 127
CURE79 127
CURE80 127
CURE81 127
CURE82 127
CURE83 127
CURE84 127
CURE85 127
CURE86 127
CURE87 127
CURE88 111
CURE89 111
CURE90 111
CURE91 111
CURE92 111
CURE93 111
CURE94 111
---------------
------------------------------------- CURE3Q=2
--------------------------------------
Variable N
---------------
CURE3Q 131
CURE1 131
CURE2Q 131
CURE4 131
CURE5 131
CURE6 131
CURE7Q 131
CURE8 131
CURE9Q 131
CURE10 131
CURE11 131
CURE12 131
CURE13 131
CURE14 131
CURE15 131
CURE16 131
CURE17 131
CURE18 131
CURE19 131
CURE20 131
CURE21 131
CURE22 131
CURE23 131
CURE24 131
CURE25 131
CURE26 131
CURE27 131
CURE28 131
CURE29Q 131
CURE30 131
CURE31 131
CURE32 131
CURE33 131
CURE34 131
CURE35 131
CURE36 131
CURE37 131
CURE38 131
CURE39 131
CURE40 131
CURE41 131
CURE42 131
CURE43Q 131
CURE44 131
CURE45 131
CURE46 131
CURE47 131
CURE48 131
CURE49 131
CURE50 131
CURE51 131
CURE52 131
CURE53 131
CURE54 131
CURE55 131
CURE56 131
CURE57 131
CURE58 131
CURE59 131
CURE60 131
CURE61 131
CURE62 131
CURE63 131
CURE64 131
CURE65 131
CURE66 131
CURE67 131
CURE68 131
CURE69 131
CURE70 131
CURE71 131
CURE72 131
CURE73 131
CURE74 131
CURE75 131
CURE76 131
CURE77 131
CURE78 131
CURE79 131
CURE80 131
CURE81 131
CURE82 131
CURE83 131
CURE84 131
CURE85 131
CURE86 131
CURE87 131
CURE88 128
CURE89 128
CURE90 128
CURE91 128
CURE92 128
CURE93 128
CURE94 128
---------------
Логит-регрессия
для двух групп CURE3Q в БД BAZA3 по п.96 Программы 4
НЦ BIOSTATISTICA.
Skype: leo_1911 E-mail: leo.biostat@gmail.com
Web-Site BIOMETRICA http:\\www.biometrica-tomsk.ru\ Tel: 8-989-830-83-10
The
LOGISTIC Procedure
Model Information
Data Set WORK.BAZA3
Response
Variable CURE3Q
Number of
Response Levels 2
Model binary logit
Optimization
Technique Fisher's scoring
Number of Observations Read 258
Number of Observations Used 229
Response Profile
Ordered Total
Value CURE3Q Frequency
1 1 101
2 2 128
Probability modeled is CURE3Q=1.
NOTE: 29 observations were deleted due to missing values for the
response or
explanatory variables.
Stepwise Selection Procedure
Step 0. Intercept
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
-2
Log L = 314.271
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -0.2369 0.1331 3.1686 0.0751 Intercept: CURE3Q=1
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
229.0000 66 <.0001
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 5.5690 0.0183
CURE2Q 1 0.0285 0.8659
CURE4 1 0.1533 0.6954
CURE5 1 0.3822 0.5364
CURE6 1 4.3555 0.0369
CURE7Q 1 3.2515 0.0714
CURE8 1 22.4786 <.0001
CURE9Q 1 0.1884 0.6642
CURE10 1 1.1305 0.2877
CURE11 1 37.2083 <.0001
CURE12 1 14.5963 0.0001
CURE13 1 3.0432 0.0811
CURE14 1 58.7885 <.0001
CURE15 1 7.4708 0.0063
CURE16 1 18.5933 <.0001
CURE17 1 27.2273 <.0001
CURE18 1 2.4524 0.1173
CURE19 1 2.0371 0.1535
CURE20 1 18.7923 <.0001
CURE21 1 0.3801 0.5376
CURE22 1 0.1154 0.7341
CURE23 1 0.2547 0.6138
CURE24 1 0.1109 0.7392
CURE25 1 33.3242 <.0001
CURE26 1 33.3429 <.0001
CURE27 1 40.5505 <.0001
CURE28 1 26.3914 <.0001
CURE29Q 1 0.0084 0.9268
CURE30 1 5.7250 0.0167
CURE31 1 10.2564 0.0014
CURE32 1 15.9611 <.0001
CURE33 1 31.4081 <.0001
CURE34 1 4.3836 0.0363
CURE35 1 12.7870 0.0003
CURE36 1 11.9851 0.0005
CURE37 1 0.0466 0.8290
CURE38 1 2.1464 0.1429
CURE39 1 3.5443 0.0598
CURE40 1 20.0213 <.0001
CURE41 1 5.2131 0.0224
CURE42 1 0.0431 0.8356
CURE43Q 1 0.0590 0.8080
CURE44 1 4.4778 0.0343
CURE45 1 4.5775 0.0324
CURE46 1 20.2262 <.0001
CURE47 1 29.2068 <.0001
CURE48 1 2.6779 0.1018
CURE49 1 3.8097 0.0510
CURE50 1 1.0903 0.2964
CURE51 1 3.0959 0.0785
CURE52 1 0.0004 0.9842
CURE53 1 2.1489 0.1427
CURE54 1 0.0201 0.8873
CURE55 1 5.1650 0.0230
CURE56 1 6.7675 0.0093
CURE57 1 0.0007 0.9795
CURE58 1 0.0500 0.8231
CURE59 1 4.1672 0.0412
CURE60 1 1.5465 0.2136
CURE61 1 8.0008 0.0047
CURE62 1 6.3368 0.0118
CURE63 1 4.5652 0.0326
CURE64 1 3.3211 0.0684
CURE65 1 2.8441 0.0917
CURE66 1 5.1827 0.0228
CURE67 1 5.7202 0.0168
CURE68 1 4.5337 0.0332
CURE69 1 5.5243 0.0188
CURE70 1 9.2876 0.0023
CURE71 1 5.1185 0.0237
CURE72 1 7.3638 0.0067
CURE73 1 7.7408 0.0054
CURE74 1 1.3607 0.2434
CURE75 1 4.0708 0.0436
CURE76 1 18.5218 <.0001
CURE77 1 1.6872 0.1940
CURE78 1 0.6832 0.4085
CURE79 1 1.4828 0.2233
CURE80 1 5.3307 0.0210
CURE81 1 0.1146 0.7349
CURE82 1 3.1639 0.0753
CURE83 1 3.3932 0.0655
CURE84 1 4.0708 0.0436
CURE85 1 0.7727 0.3794
CURE86 1 5.3307 0.0210
CURE87 1 11.3152 0.0008
CURE88 1 2.2924 0.1300
CURE89 1 4.3982 0.0360
CURE90 1 3.5596 0.0592
CURE91 1 13.5887 0.0002
CURE92 1 4.3784 0.0364
CURE93 1 4.5306 0.0333
CURE94 1 3.9709 0.0463
Step 1. Effect CURE14
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 248.548
SC 319.704 255.416
-2 Log
L 314.271 244.548
R-Square 0.2625 Max-rescaled R-Square 0.3516
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 69.7224 1 <.0001
Score 58.7885 1 <.0001
Wald 43.7972 1 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -7.9635 1.1599 47.1399 <.0001 Intercept: CURE3Q=1
CURE14 1 0.1002 0.0151 43.7972 <.0001 0.8146
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.105 1.073 1.139
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 78.1 Somers' D 0.582
Percent
Discordant 19.9 Gamma 0.594
Percent
Tied 2.0 Tau-a 0.288
Pairs 12928 c 0.791
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
218.7920 64 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 43.7972 <.0001
NOTE: No effects
for the model in Step 1 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 7.9650 0.0048
CURE2Q 1 7.7338 0.0054
CURE4 1 0.0111 0.9161
CURE5 1 0.2355 0.6274
CURE6 1 0.3573 0.5500
CURE7Q 1 0.5545 0.4565
CURE8 1 5.2971 0.0214
CURE9Q 1 1.2455 0.2644
CURE10 1 0.0801 0.7771
CURE11 1 11.4290 0.0007
CURE12 1 4.3089 0.0379
CURE13 1 2.4638 0.1165
CURE15 1 0.8440 0.3582
CURE16 1 8.7946 0.0030
CURE17 1 7.6083 0.0058
CURE18 1 26.8854 <.0001
CURE19 1 1.0479 0.3060
CURE20 1 4.8362 0.0279
CURE21 1 0.4437 0.5054
CURE22 1 9.6137 0.0019
CURE23 1 0.4585 0.4983
CURE24 1 1.7427 0.1868
CURE25 1 10.8097 0.0010
CURE26 1 9.9379 0.0016
CURE27 1 10.9724 0.0009
CURE28 1 9.4913 0.0021
CURE29Q 1 5.8331 0.0157
CURE30 1 0.4399 0.5072
CURE31 1 0.1231 0.7257
CURE32 1 0.7642 0.3820
CURE33 1 7.5518 0.0060
CURE34 1 0.3312 0.5649
CURE35 1 0.0885 0.7661
CURE36 1 1.1846 0.2764
CURE37 1 10.9818 0.0009
CURE38 1 24.4758 <.0001
CURE39 1 2.3088 0.1286
CURE40 1 0.0132 0.9085
CURE41 1 0.1017 0.7498
CURE42 1 1.7157 0.1902
CURE43Q 1 1.2108 0.2712
CURE44 1 0.7447 0.3882
CURE45 1 1.4618 0.2266
CURE46 1 6.1574 0.0131
CURE47 1 13.9721 0.0002
CURE48 1 9.0430 0.0026
CURE49 1 9.3020 0.0023
CURE50 1 0.1707 0.6795
CURE51 1 0.1626 0.6867
CURE52 1 6.5528 0.0105
CURE53 1 1.0826 0.2981
CURE54 1 3.0769 0.0794
CURE55 1 8.1227 0.0044
CURE56 1 1.2085 0.2716
CURE57 1 7.3556 0.0067
CURE58 1 4.1944 0.0406
CURE59 1 0.9939 0.3188
CURE60 1 1.3732 0.2413
CURE61 1 14.2012 0.0002
CURE62 1 4.7832 0.0287
CURE63 1 0.3912 0.5317
CURE64 1 0.4758 0.4903
CURE65 1 5.8394 0.0157
CURE66 1 14.7568 0.0001
CURE67 1 8.6912 0.0032
CURE68 1 0.0914 0.7624
CURE69 1 4.9907 0.0255
CURE70 1 15.2366 <.0001
CURE71 1 4.6051 0.0319
CURE72 1 4.9628 0.0259
CURE73 1 2.8315 0.0924
CURE74 1 1.1653 0.2804
CURE75 1 5.0900 0.0241
CURE76 1 15.1476 <.0001
CURE77 1 0.6130 0.4337
CURE78 1 0.0183 0.8925
CURE79 1 0.0048 0.9447
CURE80 1 0.0987 0.7534
CURE81 1 0.1519 0.6968
CURE82 1 0.7216 0.3956
CURE83 1 0.0079 0.9290
CURE84 1 5.0900 0.0241
CURE85 1 0.0919 0.7618
CURE86 1 0.0987 0.7534
CURE87 1 2.3967 0.1216
CURE88 1 2.9731 0.0847
CURE89 1 2.7520 0.0971
CURE90 1 3.6235 0.0570
CURE91 1 18.3440 <.0001
CURE92 1 3.5494 0.0596
CURE93 1 4.0024 0.0454
CURE94 1 4.7169 0.0299
Step 2. Effect CURE18
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 223.862
SC 319.704 234.163
-2 Log
L 314.271 217.862
R-Square 0.3436 Max-rescaled R-Square 0.4603
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 96.4088 2 <.0001
Score 81.5597 2 <.0001
Wald 57.4472 2 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -19.3529 2.7296 50.2680 <.0001 Intercept:
CURE3Q=1
CURE14 1 0.1550 0.0205 57.3250 <.0001 1.2599
CURE18 1 0.0899 0.0185 23.6420 <.0001 0.5749
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.168 1.122 1.215
CURE18 1.094 1.055 1.134
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 83.8 Somers' D 0.676
Percent
Discordant 16.2 Gamma 0.676
Percent
Tied 0.0 Tau-a 0.335
Pairs 12928 c 0.838
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
228.0320 63 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 57.3250 <.0001
CURE18 1 23.6420 <.0001
NOTE: No effects for the model in Step
2 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 0.1607 0.6885
CURE2Q 1 9.0783 0.0026
CURE4 1 0.5505 0.4581
CURE5 1 0.0043 0.9477
CURE6 1 1.2611 0.2615
CURE7Q 1 1.9305 0.1647
CURE8 1 2.0456 0.1526
CURE9Q 1 0.5318 0.4659
CURE10 1 0.1177 0.7315
CURE11 1 5.2700 0.0217
CURE12 1 4.1790 0.0409
CURE13 1 0.0026 0.9590
CURE15 1 0.0249 0.8747
CURE16 1 3.6803 0.0551
CURE17 1 4.4442 0.0350
CURE19 1 0.7842 0.3759
CURE20 1 6.2286 0.0126
CURE21 1 0.2620 0.6088
CURE22 1 11.2030 0.0008
CURE23 1 1.1759 0.2782
CURE24 1 4.1048 0.0428
CURE25 1 4.8324 0.0279
CURE26 1 6.2804 0.0122
CURE27 1 4.7796 0.0288
CURE28 1 7.3376 0.0068
CURE29Q 1 8.0116 0.0046
CURE30 1 2.9240 0.0873
CURE31 1 4.5603 0.0327
CURE32 1 0.0091 0.9240
CURE33 1 3.8494 0.0498
CURE34 1 1.0522 0.3050
CURE35 1 3.2488 0.0715
CURE36 1 0.1380 0.7103
CURE37 1 9.6389 0.0019
CURE38 1 19.6677 <.0001
CURE39 1 3.2325 0.0722
CURE40 1 0.0075 0.9309
CURE41 1 1.1047 0.2932
CURE42 1 1.4416 0.2299
CURE43Q 1 0.2153 0.6427
CURE44 1 1.7860 0.1814
CURE45 1 2.4257 0.1194
CURE46 1 0.2223 0.6373
CURE47 1 6.1207 0.0134
CURE48 1 22.9430 <.0001
CURE49 1 7.0794 0.0078
CURE50 1 0.3869 0.5339
CURE51 1 0.1313 0.7170
CURE52 1 7.8825 0.0050
CURE53 1 0.0954 0.7574
CURE54 1 1.7320 0.1882
CURE55 1 12.0794 0.0005
CURE56 1 0.0208 0.8853
CURE57 1 8.2752 0.0040
CURE58 1 1.9715 0.1603
CURE59 1 0.7024 0.4020
CURE60 1 0.8650 0.3523
CURE61 1 16.3447 <.0001
CURE62 1 2.1055 0.1468
CURE63 1 1.1751 0.2783
CURE64 1 3.9718 0.0463
CURE65 1 21.7391 <.0001
CURE66 1 10.0112 0.0016
CURE67 1 11.7928 0.0006
CURE68 1 1.8419 0.1747
CURE69 1 1.4397 0.2302
CURE70 1 12.6466 0.0004
CURE71 1 3.1852 0.0743
CURE72 1 8.5962 0.0034
CURE73 1 10.0883 0.0015
CURE74 1 1.5448 0.2139
CURE75 1 2.4152 0.1202
CURE76 1 21.4201 <.0001
CURE77 1 0.0049 0.9443
CURE78 1 0.8556 0.3550
CURE79 1 0.2308 0.6309
CURE80 1 1.1468 0.2842
CURE81 1 0.0248 0.8748
CURE82 1 0.2487 0.6180
CURE83 1 0.9442 0.3312
CURE84 1 2.4152 0.1202
CURE85 1 0.1334 0.7149
CURE86 1 1.1468 0.2842
CURE87 1 2.7019 0.1002
CURE88 1 4.8092 0.0283
CURE89 1 5.4102 0.0200
CURE90 1 6.7677 0.0093
CURE91 1 29.8677 <.0001
CURE92 1 7.1494 0.0075
CURE93 1 7.9015 0.0049
CURE94 1 10.1929 0.0014
Step 3. Effect CURE91
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 189.036
SC 319.704 202.771
-2 Log
L 314.271 181.036
R-Square 0.4411 Max-rescaled R-Square 0.5909
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 133.2346 3 <.0001
Score 100.0033 3 <.0001
Wald 60.4603 3 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -24.4450 3.3722 52.5461 <.0001 Intercept:
CURE3Q=1
CURE14 1 0.1916 0.0260 54.3341 <.0001 1.5571
CURE18 1 0.1293 0.0226 32.7161 <.0001 0.8268
CURE91 1 -1.0064 0.2682 14.0842 0.0002 -1.0061
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.211 1.151 1.274
CURE18 1.138 1.089 1.190
CURE91 0.366 0.216 0.618
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 88.7 Somers' D 0.773
Percent
Discordant 11.3 Gamma 0.773
Percent
Tied 0.0 Tau-a 0.383
Pairs 12928 c 0.887
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
203.6141 62 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 54.3341 <.0001
CURE18 1 32.7161 <.0001
CURE91 1 14.0842 0.0002
NOTE: No effects
for the model in Step 3 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 0.1461 0.7023
CURE2Q 1 8.4693 0.0036
CURE4 1 0.0463 0.8297
CURE5 1 0.0074 0.9314
CURE6 1 3.3461 0.0674
CURE7Q 1 3.9517 0.0468
CURE8 1 3.8524 0.0497
CURE9Q 1 0.0000 0.9964
CURE10 1 0.0144 0.9044
CURE11 1 11.9708 0.0005
CURE12 1 7.9281 0.0049
CURE13 1 0.0268 0.8699
CURE15 1 0.1030 0.7483
CURE16 1 2.3327 0.1267
CURE17 1 3.9270 0.0475
CURE19 1 0.0310 0.8603
CURE20 1 13.8209 0.0002
CURE21 1 0.3235 0.5695
CURE22 1 8.7642 0.0031
CURE23 1 0.6753 0.4112
CURE24 1 4.7673 0.0290
CURE25 1 1.1792 0.2775
CURE26 1 3.5434 0.0598
CURE27 1 0.9931 0.3190
CURE28 1 1.6276 0.2020
CURE29Q 1 6.5999 0.0102
CURE30 1 11.9000 0.0006
CURE31 1 14.2195 0.0002
CURE32 1 1.3799 0.2401
CURE33 1 1.2333 0.2668
CURE34 1 0.4044 0.5248
CURE35 1 8.6643 0.0032
CURE36 1 1.4849 0.2230
CURE37 1 7.6668 0.0056
CURE38 1 7.2344 0.0072
CURE39 1 5.2941 0.0214
CURE40 1 0.0202 0.8871
CURE41 1 2.3458 0.1256
CURE42 1 2.8769 0.0899
CURE43Q 1 0.1971 0.6570
CURE44 1 2.9316 0.0869
CURE45 1 2.8303 0.0925
CURE46 1 0.1458 0.7025
CURE47 1 9.8355 0.0017
CURE48 1 18.3824 <.0001
CURE49 1 2.2301 0.1353
CURE50 1 2.0481 0.1524
CURE51 1 0.0000 0.9984
CURE52 1 8.1757 0.0042
CURE53 1 0.0001 0.9934
CURE54 1 0.4683 0.4938
CURE55 1 11.9377 0.0006
CURE56 1 1.4015 0.2365
CURE57 1 10.2376 0.0014
CURE58 1 7.3984 0.0065
CURE59 1 0.6331 0.4262
CURE60 1 2.0740 0.1498
CURE61 1 14.2334 0.0002
CURE62 1 3.8356 0.0502
CURE63 1 0.4830 0.4871
CURE64 1 3.9875 0.0458
CURE65 1 20.9053 <.0001
CURE66 1 15.3681 <.0001
CURE67 1 8.2040 0.0042
CURE68 1 0.6627 0.4156
CURE69 1 0.0738 0.7858
CURE70 1 5.5318 0.0187
CURE71 1 3.0967 0.0785
CURE72 1 9.1785 0.0024
CURE73 1 8.3043 0.0040
CURE74 1 2.4287 0.1191
CURE75 1 0.4507 0.5020
CURE76 1 11.5271 0.0007
CURE77 1 0.0003 0.9869
CURE78 1 1.7377 0.1874
CURE79 1 0.0159 0.8997
CURE80 1 0.1735 0.6770
CURE81 1 0.0493 0.8243
CURE82 1 0.4136 0.5202
CURE83 1 0.0171 0.8959
CURE84 1 0.4507 0.5020
CURE85 1 0.0320 0.8581
CURE86 1 0.1735 0.6770
CURE87 1 1.0579 0.3037
CURE88 1 4.4206 0.0355
CURE89 1 3.7541 0.0527
CURE90 1 4.6991 0.0302
CURE92 1 5.1792 0.0229
CURE93 1 5.3351 0.0209
CURE94 1 6.2702 0.0123
Step 4. Effect CURE65
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 169.205
SC 319.704 186.373
-2 Log
L 314.271 159.205
R-Square 0.4919 Max-rescaled R-Square 0.6590
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 155.0660 4 <.0001
Score 111.6233 4 <.0001
Wald 55.5679 4 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error Chi-Square
Pr > ChiSq Estimate Label
Intercept 1 -286.9 61.7970 21.5499 <.0001 Intercept: CURE3Q=1
CURE14 1 0.2463 0.0340 52.5614 <.0001 2.0025
CURE18 1 0.1891 0.0305 38.4556 <.0001 1.2087
CURE65 1 8.5179 1.9726 18.6464 <.0001 0.6004
CURE91 1 -1.0709 0.2989 12.8367 0.0003 -1.0706
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.279 1.197 1.367
CURE18 1.208 1.138 1.283
CURE65 >999.999 104.763 >999.999
CURE91 0.343 0.191 0.616
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 92.9 Somers' D 0.858
Percent
Discordant 7.1 Gamma 0.858
Percent
Tied 0.0 Tau-a 0.425
Pairs 12928 c 0.929
Residual
Chi-Square Test
Chi-Square DF Pr > ChiSq
223.6092 61 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 52.5614 <.0001
CURE18 1 38.4556 <.0001
CURE65 1 18.6464 <.0001
CURE91 1 12.8367 0.0003
NOTE: No effects
for the model in Step 4 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 0.0019 0.9649
CURE2Q 1 10.8521 0.0010
CURE4 1 0.6304 0.4272
CURE5 1 0.1738 0.6768
CURE6 1 4.8722 0.0273
CURE7Q 1 3.9617 0.0465
CURE8 1 5.0613 0.0245
CURE9Q 1 0.1411 0.7072
CURE10 1 0.6170 0.4322
CURE11 1 10.1925 0.0014
CURE12 1 4.5721 0.0325
CURE13 1 0.0445 0.8330
CURE15 1 0.5651 0.4522
CURE16 1 1.1164 0.2907
CURE17 1 0.6847 0.4080
CURE19 1 0.1886 0.6640
CURE20 1 8.9474 0.0028
CURE21 1 0.0008 0.9780
CURE22 1 6.5708 0.0104
CURE23 1 0.1799 0.6715
CURE24 1 4.3391 0.0372
CURE25 1 1.1893 0.2755
CURE26 1 4.5203 0.0335
CURE27 1 0.1691 0.6809
CURE28 1 0.1567 0.6922
CURE29Q 1 6.2791 0.0122
CURE30 1 35.8806 <.0001
CURE31 1 22.5251 <.0001
CURE32 1 8.7285 0.0031
CURE33 1 0.0000 0.9991
CURE34 1 0.3610 0.5479
CURE35 1 5.8137 0.0159
CURE36 1 0.0384 0.8447
CURE37 1 8.2516 0.0041
CURE38 1 8.3678 0.0038
CURE39 1 6.6896 0.0097
CURE40 1 0.6541 0.4186
CURE41 1 3.6978 0.0545
CURE42 1 4.6919 0.0303
CURE43Q 1 0.8458 0.3577
CURE44 1 2.9553 0.0856
CURE45 1 2.0148 0.1558
CURE46 1 4.1345 0.0420
CURE47 1 6.5636 0.0104
CURE48 1 18.3933 <.0001
CURE49 1 7.3530 0.0067
CURE50 1 0.3748 0.5404
CURE51 1 0.0648 0.7991
CURE52 1 9.8276 0.0017
CURE53 1 0.3564 0.5505
CURE54 1 0.8760 0.3493
CURE55 1 6.4085 0.0114
CURE56 1 2.5217 0.1123
CURE57 1 7.5410 0.0060
CURE58 1 9.2368 0.0024
CURE59 1 0.0155 0.9008
CURE60 1 0.1341 0.7143
CURE61 1 10.9388 0.0009
CURE62 1 9.0735 0.0026
CURE63 1 0.4376 0.5083
CURE64 1 0.4357 0.5092
CURE66 1 18.3344 <.0001
CURE67 1 6.7393 0.0094
CURE68 1 2.5965 0.1071
CURE69 1 1.2744 0.2589
CURE70 1 2.3873 0.1223
CURE71 1 0.6817 0.4090
CURE72 1 9.6186 0.0019
CURE73 1 1.1422 0.2852
CURE74 1 9.0854 0.0026
CURE75 1 0.0061 0.9379
CURE76 1 16.3416 <.0001
CURE77 1 0.0065 0.9357
CURE78 1 2.1295 0.1445
CURE79 1 0.0003 0.9855
CURE80 1 0.0435 0.8348
CURE81 1 0.1482 0.7002
CURE82 1 0.2782 0.5979
CURE83 1 0.0179 0.8935
CURE84 1 0.0061 0.9379
CURE85 1 0.0046 0.9458
CURE86 1 0.0435 0.8348
CURE87 1 1.1978 0.2738
CURE88 1 3.8909 0.0485
CURE89 1 3.3996 0.0652
CURE90 1 3.9751 0.0462
CURE92 1 4.3999 0.0359
CURE93 1 4.6431 0.0312
CURE94 1 6.2599 0.0124
Step 5. Effect CURE30
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 146.833
SC 319.704 167.435
-2 Log
L 314.271 134.833
R-Square 0.5432 Max-rescaled R-Square 0.7277
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 179.4377 5 <.0001
Score 117.7354 5 <.0001
Wald 46.6414 5 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -428.4 78.7351 29.6053 <.0001 Intercept: CURE3Q=1
CURE14 1 0.3763 0.0566 44.1877 <.0001 3.0585
CURE18 1 0.2735 0.0431 40.2565 <.0001 1.7487
CURE30 1 -0.0848 0.0187 20.6568 <.0001 -0.5914
CURE65 1 12.8402 2.4690 27.0467 <.0001 0.9050
CURE91 1 -1.4398 0.3885 13.7378 0.0002 -1.4394
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.457 1.304 1.628
CURE18 1.315 1.208 1.431
CURE30 0.919 0.886 0.953
CURE65 >999.999 >999.999 >999.999
CURE91 0.237 0.111 0.507
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 93.9 Somers' D 0.879
Percent
Discordant 6.1 Gamma 0.879
Percent
Tied 0.0 Tau-a 0.435
Pairs 12928 c 0.939
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
144.4682 60 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 44.1877 <.0001
CURE18 1 40.2565 <.0001
CURE30 1 20.6568 <.0001
CURE65 1 27.0467 <.0001
CURE91 1 13.7378 0.0002
NOTE: No effects
for the model in Step 5 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 3.4761 0.0623
CURE2Q 1 5.9874 0.0144
CURE4 1 0.4686 0.4936
CURE5 1 2.4063 0.1208
CURE6 1 14.1879 0.0002
CURE7Q 1 13.4594 0.0002
CURE8 1 18.0813 <.0001
CURE9Q 1 3.6984 0.0545
CURE10 1 5.3583 0.0206
CURE11 1 23.4229 <.0001
CURE12 1 16.9440 <.0001
CURE13 1 4.1500 0.0416
CURE15 1 1.4970 0.2211
CURE16 1 1.1569 0.2821
CURE17 1 1.5152 0.2183
CURE19 1 3.3625 0.0667
CURE20 1 4.6110 0.0318
CURE21 1 0.0346 0.8525
CURE22 1 2.5817 0.1081
CURE23 1 1.4740 0.2247
CURE24 1 5.2271 0.0222
CURE25 1 6.6966 0.0097
CURE26 1 13.8661 0.0002
CURE27 1 12.8232 0.0003
CURE28 1 8.8897 0.0029
CURE29Q 1 2.0801 0.1492
CURE31 1 0.6314 0.4268
CURE32 1 12.7693 0.0004
CURE33 1 35.8102 <.0001
CURE34 1 6.5821 0.0103
CURE35 1 1.9584 0.1617
CURE36 1 1.3244 0.2498
CURE37 1 7.9423 0.0048
CURE38 1 2.7482 0.0974
CURE39 1 0.2186 0.6401
CURE40 1 1.2509 0.2634
CURE41 1 0.8578 0.3544
CURE42 1 0.0970 0.7554
CURE43Q 1 0.0548 0.8149
CURE44 1 0.2989 0.5846
CURE45 1 0.6004 0.4384
CURE46 1 10.6528 0.0011
CURE47 1 15.0622 0.0001
CURE48 1 27.1634 <.0001
CURE49 1 4.9249 0.0265
CURE50 1 3.5259 0.0604
CURE51 1 0.9827 0.3215
CURE52 1 6.4650 0.0110
CURE53 1 2.7240 0.0988
CURE54 1 1.3402 0.2470
CURE55 1 5.3100 0.0212
CURE56 1 3.2210 0.0727
CURE57 1 17.0598 <.0001
CURE58 1 5.2007 0.0226
CURE59 1 1.7381 0.1874
CURE60 1 0.0081 0.9284
CURE61 1 7.4630 0.0063
CURE62 1 10.5638 0.0012
CURE63 1 0.0521 0.8195
CURE64 1 0.0404 0.8408
CURE66 1 14.3978 0.0001
CURE67 1 0.1002 0.7516
CURE68 1 0.0338 0.8540
CURE69 1 4.2325 0.0397
CURE70 1 1.1098 0.2921
CURE71 1 2.0807 0.1492
CURE72 1 4.1837 0.0408
CURE73 1 1.4813 0.2236
CURE74 1 5.3489 0.0207
CURE75 1 0.0402 0.8411
CURE76 1 15.8747 <.0001
CURE77 1 0.0147 0.9035
CURE78 1 2.8129 0.0935
CURE79 1 0.1034 0.7478
CURE80 1 2.2041 0.1376
CURE81 1 1.2791 0.2581
CURE82 1 0.0141 0.9054
CURE83 1 0.0735 0.7863
CURE84 1 0.0402 0.8411
CURE85 1 0.0847 0.7710
CURE86 1 2.2041 0.1376
CURE87 1 0.0974 0.7549
CURE88 1 8.5124 0.0035
CURE89 1 6.9868 0.0082
CURE90 1 8.4965 0.0036
CURE92 1 9.2242 0.0024
CURE93 1 9.5460 0.0020
CURE94 1 12.1349 0.0005
Step 6. Effect CURE33
entered:
Model Convergence Status
Convergence
criterion (GCONV=1E-8) satisfied.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 102.964
SC 319.704 127.000
-2 Log
L 314.271 88.964
R-Square 0.6261 Max-rescaled R-Square 0.8388
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 225.3064 6 <.0001
Score 129.8003 6 <.0001
Wald 34.8427 6 <.0001
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -601.5 109.4 30.2359 <.0001 Intercept: CURE3Q=1
CURE14 1 0.5215 0.0926 31.7407 <.0001 4.2393
CURE18 1 0.4009 0.0732 29.9833 <.0001 2.5629
CURE30 1 -0.3813 0.0896 18.1161 <.0001 -2.6593
CURE33 1 11.9709 2.5867 21.4172 <.0001 2.4998
CURE65 1 16.4142 3.1242 27.6039 <.0001 1.1569
CURE91 1 -3.2064 0.7472 18.4139 <.0001 -3.2055
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 1.685 1.405 2.020
CURE18 1.493 1.294 1.724
CURE30 0.683 0.573 0.814
CURE33 >999.999 993.341 >999.999
CURE65 >999.999 >999.999 >999.999
CURE91 0.041 0.009 0.175
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 97.4 Somers' D 0.948
Percent
Discordant 2.6 Gamma 0.948
Percent
Tied 0.0 Tau-a 0.470
Pairs 12928 c 0.974
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
132.5130 57 <.0001
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 31.7407 <.0001
CURE18 1 29.9833 <.0001
CURE30 1 18.1161 <.0001
CURE33 1 21.4172 <.0001
CURE65 1 27.6039 <.0001
CURE91 1 18.4139 <.0001
NOTE: No effects
for the model in Step 6 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 0.1257 0.7230
CURE2Q 1 17.5428 <.0001
CURE4 1 0.1059 0.7448
CURE5 1 0.4251 0.5144
CURE6 1 8.9158 0.0028
CURE7Q 1 13.8756 0.0002
CURE8 1 10.8204 0.0010
CURE9Q 1 1.9198 0.1659
CURE10 1 4.5475 0.0330
CURE11 1 19.7479 <.0001
CURE12 1 9.5884 0.0020
CURE13 1 0.0787 0.7791
CURE15 1 0.7110 0.3991
CURE16 1 0.0001 0.9927
CURE17 1 0.1268 0.7218
CURE19 1 4.9678 0.0258
CURE20 1 6.1024 0.0135
CURE21 1 0.1573 0.6917
CURE22 1 0.6416 0.4231
CURE23 1 2.3825 0.1227
CURE24 1 8.4889 0.0036
CURE25 1 0.7384 0.3902
CURE26 1 3.8973 0.0484
CURE27 1 0.4509 0.5019
CURE28 1 0.0295 0.8636
CURE29Q 1 13.4162 0.0002
CURE31 1 0.3540 0.5519
CURE32 1 0.0373 0.8468
CURE34 1 3.4801 0.0621
CURE35 1 0.1477 0.7008
CURE36 1 1.2367 0.2661
CURE37 1 32.3752 <.0001
CURE38 1 15.3674 <.0001
CURE39 1 17.6767 <.0001
CURE40 1 1.2474 0.2640
CURE41 1 5.8139 0.0159
CURE42 1 8.3482 0.0039
CURE43Q 1 3.6090 0.0575
CURE44 1 0.0407 0.8401
CURE45 1 0.5457 0.4601
CURE46 1 10.9818 0.0009
CURE47 1 7.0000 0.0082
CURE48 1 19.3211 <.0001
CURE49 1 5.1647 0.0231
CURE50 1 1.5294 0.2162
CURE51 1 1.5569 0.2121
CURE52 1 14.3353 0.0002
CURE53 1 0.2668 0.6055
CURE54 1 0.0691 0.7926
CURE55 1 4.7486 0.0293
CURE56 1 6.4583 0.0110
CURE57 1 15.4812 <.0001
CURE58 1 1.9775 0.1596
CURE59 1 0.7409 0.3894
CURE60 1 0.1257 0.7229
CURE61 1 6.1570 0.0131
CURE62 1 6.3569 0.0117
CURE63 1 0.0716 0.7890
CURE64 1 2.2527 0.1334
CURE66 1 2.6860 0.1012
CURE67 1 4.2108 0.0402
CURE68 1 1.2856 0.2569
CURE69 1 6.5523 0.0105
CURE70 1 0.2740 0.6007
CURE71 1 3.8680 0.0492
CURE72 1 4.5826 0.0323
CURE73 1 1.3364 0.2477
CURE74 1 10.3337 0.0013
CURE75 1 0.9856 0.3208
CURE76 1 8.3174 0.0039
CURE77 1 1.1207 0.2898
CURE78 1 8.3633 0.0038
CURE79 1 0.0449 0.8321
CURE80 1 0.0170 0.8963
CURE81 1 0.8038 0.3700
CURE82 1 1.2516 0.2633
CURE83 1 0.1653 0.6843
CURE84 1 0.9856 0.3208
CURE85 1 0.0562 0.8126
CURE86 1 0.0170 0.8963
CURE87 1 0.0007 0.9795
CURE88 1 3.8816 0.0488
CURE89 1 3.0153 0.0825
CURE90 1 3.7809 0.0518
CURE92 1 4.4178 0.0356
CURE93 1 4.5429 0.0331
CURE94 1 5.8155 0.0159
Step 7. Effect CURE37
entered:
Model Convergence Status
Complete
separation of data points detected.
WARNING: The maximum
likelihood estimate does not exist.
WARNING: The LOGISTIC procedure continues in spite of the above warning.
Results
shown are based on the
last maximum likelihood iteration. Validity of the
model fit is
questionable.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 30.449
SC 319.704 57.919
-2 Log
L 314.271 14.449
R-Square 0.7300 Max-rescaled R-Square 0.9779
Testing Global Null Hypothesis:
BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 299.8212 7 <.0001
Score 143.6515 7 <.0001
Wald 9.5204 7 0.2174
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -2800.3 1255.4 4.9753 0.0257 Intercept: CURE3Q=1
CURE14 1 4.8504 2.3578 4.2318 0.0397 39.4270
CURE18 1 2.4127 1.1373 4.5005 0.0339 15.4244
CURE30 1 -4.0578 1.9316 4.4133 0.0357 -28.3025
CURE33 1 148.0 71.0327 4.3434 0.0372 30.9131
CURE37 1 -6.5692 3.2979 3.9677 0.0464 -18.1285
CURE65 1 62.4760 27.2094 5.2722 0.0217 4.4035
CURE91 1 -24.6806 11.5044 4.6024 0.0319 -24.6731
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 127.788 1.258 >999.999
CURE18 11.164 1.202 103.732
CURE30 0.017 <0.001 0.762
CURE33 >999.999 >999.999 >999.999
CURE37 0.001 <0.001 0.900
CURE65 >999.999 >999.999 >999.999
CURE91 <0.001 <0.001 0.119
Association of Predicted
Probabilities and Observed Responses
Percent
Concordant 100.0 Somers' D 1.000
Percent
Discordant 0.0 Gamma 1.000
Percent
Tied 0.0 Tau-a 0.495
Pairs 12928 c 1.000
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
9.1037 15 0.8720
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 4.2318 0.0397
CURE18 1 4.5005 0.0339
CURE30 1 4.4133 0.0357
CURE33 1 4.3434 0.0372
CURE37 1 3.9677 0.0464
CURE65 1 5.2722 0.0217
CURE91 1 4.6024 0.0319
NOTE: No effects
for the model in Step 7 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 8.5560 0.0034
CURE2Q 1 7.5772 0.0059
CURE4 1 8.6332 0.0033
CURE5 1 8.2595 0.0041
CURE6 1 7.9870 0.0047
CURE7Q 1 8.2990 0.0040
CURE8 1 8.9781 0.0027
CURE9Q 1 7.4230 0.0064
CURE10 1 8.9410 0.0028
CURE11 1 8.8648 0.0029
CURE12 1 8.5289 0.0035
CURE13 1 8.8755 0.0029
CURE15 1 8.5247 0.0035
CURE16 1 8.2574 0.0041
CURE17 1 7.4013 0.0065
CURE19 1 8.6284 0.0033
CURE20 1 9.0864 0.0026
CURE21 1 7.6645 0.0056
CURE22 1 7.9223 0.0049
CURE23 1 9.0900 0.0026
CURE24 1 8.9851 0.0027
CURE25 1 9.0647 0.0026
CURE26 1 9.0843 0.0026
CURE27 1 9.0853 0.0026
CURE28 1 9.0812 0.0026
CURE29Q 1 7.6922 0.0055
CURE31 1 7.4533 0.0063
CURE32 1 8.8057 0.0030
CURE34 1 7.8966 0.0050
CURE35 1 8.5311 0.0035
CURE36 1 8.9176 0.0028
CURE38 1 8.0345 0.0046
CURE39 1 8.9316 0.0028
CURE40 1 9.0521 0.0026
CURE41 1 9.0243 0.0027
CURE42 1 8.9489 0.0028
CURE43Q 1 8.8942 0.0029
CURE44 1 8.9811 0.0027
CURE45 1 9.0553 0.0026
CURE46 1 8.9555 0.0028
CURE47 1 7.5156 0.0061
CURE48 1 7.3617 0.0067
CURE49 1 8.9694 0.0027
CURE50 1 8.9945 0.0027
CURE51 1 8.3279 0.0039
CURE52 1 8.3331 0.0039
CURE53 1 8.7977 0.0030
CURE54 1 8.3598 0.0038
CURE55 1 9.0134 0.0027
CURE56 1 8.6876 0.0032
CURE57 1 7.5778 0.0059
CURE58 1 8.7388 0.0031
CURE59 1 8.7116 0.0032
CURE60 1 8.3469 0.0039
CURE61 1 8.6302 0.0033
CURE62 1 7.4392 0.0064
CURE63 1 8.4776 0.0036
CURE64 1 8.6430 0.0033
CURE66 1 8.6157 0.0033
CURE67 1 7.9291 0.0049
CURE68 1 8.6411 0.0033
CURE69 1 8.6071 0.0033
CURE70 1 8.9589 0.0028
CURE71 1 8.5162 0.0035
CURE72 1 8.8109 0.0030
CURE73 1 8.4258 0.0037
CURE74 1 9.0287 0.0027
CURE75 1 8.6607 0.0033
CURE76 1 8.9810 0.0027
CURE77 1 8.5285 0.0035
CURE78 1 9.0371 0.0026
CURE79 1 9.0454 0.0026
CURE80 1 8.9168 0.0028
CURE81 1 9.0158 0.0027
CURE82 1 8.8451 0.0029
CURE83 1 8.1508 0.0043
CURE84 1 8.6607 0.0033
CURE85 1 9.0439 0.0026
CURE86 1 8.9168 0.0028
CURE87 1 8.7500 0.0031
CURE88 1 8.8753 0.0029
CURE89 1 8.5337 0.0035
CURE90 1 9.0479 0.0026
CURE92 1 9.0648 0.0026
CURE93 1 9.0724 0.0026
CURE94 1 7.3617 0.0067
Step 8. Effect CURE23
entered:
Model Convergence Status
Complete separation of data
points detected.
WARNING: The maximum
likelihood estimate does not exist.
WARNING: The LOGISTIC procedure continues in spite of the above warning.
Results
shown are based on the
last maximum likelihood iteration. Validity of the
model fit is
questionable.
Model Fit Statistics
Intercept
Intercept and
Criterion Only Covariates
AIC 316.271 20.084
SC 319.704 50.987
-2 Log
L 314.271 2.084
R-Square 0.7442 Max-rescaled R-Square 0.9969
Testing
Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood
Ratio 312.1867 8 <.0001
Score 143.6515 8 <.0001
Wald 12.9326 8 0.1142
Analysis of
Maximum Likelihood Estimates
Standard Wald Standardized
Parameter DF Estimate Error
Chi-Square Pr > ChiSq Estimate Label
Intercept 1 -2200.6 801.2 7.5440 0.0060 Intercept: CURE3Q=1
CURE14 1 2.7723 0.8185 11.4723 0.0007 22.5351
CURE18 1 1.4316 0.4181 11.7265 0.0006 9.1523
CURE23 1 -0.1958 0.0956 4.1930 0.0406 -2.2168
CURE30 1 -2.6378 0.8691 9.2108 0.0024 -18.3982
CURE33 1 88.3448 27.7874 10.1080 0.0015 18.4481
CURE37 1 -3.3687 1.1091 9.2260 0.0024 -9.2964
CURE65 1 55.5208 22.5280 6.0739 0.0137 3.9132
CURE91 1 -15.2040 5.5996 7.3722 0.0066 -15.1994
Odds Ratio Estimates
Point 95% Wald
Effect Estimate Confidence Limits
CURE14 15.995 3.216 79.560
CURE18 4.185 1.845 9.497
CURE23 0.822 0.682 0.992
CURE30 0.072 0.013 0.393
CURE33 >999.999 >999.999 >999.999
CURE37 0.034 0.004 0.303
CURE65 >999.999 >999.999 >999.999
CURE91 <0.001 <0.001 0.015
Association of
Predicted Probabilities and Observed Responses
Percent
Concordant 100.0 Somers' D 1.000
Percent
Discordant 0.0 Gamma 1.000
Percent
Tied 0.0 Tau-a 0.495
Pairs 12928 c 1.000
Residual Chi-Square Test
Chi-Square DF Pr > ChiSq
1.0611 27 1.0000
Analysis of
Effects Eligible for Removal
Wald
Effect DF Chi-Square Pr > ChiSq
CURE14 1 11.4723 0.0007
CURE18 1 11.7265 0.0006
CURE23 1 4.1930 0.0406
CURE30 1 9.2108 0.0024
CURE33 1 10.1080 0.0015
CURE37 1 9.2260 0.0024
CURE65 1 6.0739 0.0137
CURE91 1 7.3722 0.0066
NOTE: No effects
for the model in Step 8 are removed.
Analysis of
Effects Eligible for Entry
Score
Effect DF Chi-Square Pr > ChiSq
CURE1 1 1.0282 0.3106
CURE2Q 1 1.0281 0.3106
CURE4 1 1.0390 0.3081
CURE5 1 1.0391 0.3080
CURE6 1 1.0294 0.3103
CURE7Q 1 1.0294 0.3103
CURE8 1 1.0302 0.3101
CURE9Q 1 1.0368 0.3086
CURE10 1 1.0455 0.3065
CURE11 1 1.0304 0.3101
CURE12 1 1.0367 0.3086
CURE13 1 1.0371 0.3085
CURE15 1 1.0282 0.3106
CURE16 1 1.0281 0.3106
CURE17 1 1.0333 0.3094
CURE19 1 1.0285 0.3105
CURE20 1 1.0305 0.3100
CURE21 1 1.0511 0.3053
CURE22 1 1.0420 0.3074
CURE24 1 1.0299 0.3102
CURE25 1 1.0540 0.3046
CURE26 1 1.0327 0.3095
CURE27 1 1.0426 0.3072
CURE28 1 1.0290 0.3104
CURE29Q 1 1.0294 0.3103
CURE31 1 1.0476 0.3061
CURE32 1 1.0313 0.3098
CURE34 1 1.0301 0.3101
CURE35 1 1.0315 0.3098
CURE36 1 1.0282 0.3106
CURE38 1 1.0281 0.3106
CURE39 1 1.0356 0.3089
CURE40 1 1.0419 0.3074
CURE41 1 1.0287 0.3105
CURE42 1 1.0289 0.3104
CURE43Q 1 1.0291 0.3104
CURE44 1 1.0297 0.3102
CURE45 1 1.0301 0.3101
CURE46 1 1.0291 0.3104
CURE47 1 1.0346 0.3091
CURE48 1 1.0332 0.3094
CURE49 1 1.0451 0.3066
CURE50 1 1.0427 0.3072
CURE51 1 1.0302 0.3101
CURE52 1 1.0285 0.3105
CURE53 1 1.0394 0.3080
CURE54 1 1.0382 0.3082
CURE55 1 1.0378 0.3083
CURE56 1 1.0382 0.3082
CURE57 1 1.0354 0.3089
CURE58 1 1.0292 0.3104
CURE59 1 1.0344 0.3091
CURE60 1 1.0309 0.3099
CURE61 1 1.0370 0.3085
CURE62 1 1.0283 0.3106
CURE63 1 1.0281 0.3106
CURE64 1 1.0322 0.3096
CURE66 1 1.0311 0.3099
CURE67 1 1.0285 0.3105
CURE68 1 1.0292 0.3103
CURE69 1 1.0283 0.3105
CURE70 1 1.0306 0.3100
CURE71 1 1.0283 0.3105
CURE72 1 1.0285 0.3105
CURE73 1 1.0292 0.3104
CURE74 1 1.0284 0.3105
CURE75 1 1.0321 0.3097
CURE76 1 1.0392 0.3080
CURE77 1 1.0511 0.3053
CURE78 1 1.0351 0.3090
CURE79 1 1.0282 0.3106
CURE80 1 1.0494 0.3056
CURE81 1 1.0357 0.3088
CURE82 1 1.0385 0.3082
CURE83 1 1.0281 0.3106
CURE84 1 1.0321 0.3097
CURE85 1 1.0282 0.3106
CURE86 1 1.0494 0.3056
CURE87 1 1.0372 0.3085
CURE88 1 1.0300 0.3102
CURE89 1 1.0281 0.3106
CURE90 1 1.0342 0.3092
CURE92 1 1.0372 0.3085
CURE93 1 1.0307 0.3100
CURE94 1 1.0316 0.3098
NOTE: No (additional) effects
met the 0.1 significance level for entry into the
model.
Summary of Stepwise Selection
Effect Number Score Wald
Step Entered Removed DF In Chi-Square Chi-Square Pr > ChiSq
1 CURE14 1 1 58.7885 <.0001
2 CURE18 1 2 26.8854 <.0001
3 CURE91 1 3 29.8677 <.0001
4 CURE65 1 4 20.9053 <.0001
5 CURE30 1 5 35.8806 <.0001
6 CURE33 1 6 35.8102 <.0001
7 CURE37 1 7 32.3752 <.0001
8 CURE23 1 8 9.0900 0.0026
Partition for
the Hosmer and Lemeshow Test
CURE3Q = 1 CURE3Q = 2
Group Total Observed Expected Observed Expected
1 93 0 0.00 93 93.00
2 22 0 0.11 22 21.89
3 24 11 11.03 13 12.97
4 25 25 24.92 0 0.08
5 65 65 65.00 0 0.00
Hosmer and Lemeshow Goodness-of-Fit Test
Chi-Square DF Pr > ChiSq
0.1934 3 0.9786
NOTE: In calculating the Expected values, predicted probabilities less
than 1E-6 and
greater than 0.999999 were
changed to 1E-6 and 0.999999 respectively.
Classification
Table
Correct Incorrect Percentages
Prob Non- Non- Sensi- Speci- False False
Level Event Event Event Event Correct tivity ficity POS NEG
0.000 101 0 128 0 44.1 100.0 0.0 55.9 .
0.020 101 112 16 0 93.0 100.0 87.5 13.7 0.0
0.040 101 119 9 0 96.1 100.0 93.0 8.2 0.0
0.060 101 126 2 0 99.1 100.0 98.4 1.9 0.0
0.080 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.100 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.120 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.140 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.160 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.180 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.200 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.220 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.240 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.260 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.280 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.300 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.320 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.340 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.360 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.380 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.400 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.420 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.440 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.460 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.480 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.500 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.520 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.540 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.560 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.580 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.600 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.620 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.640 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.660 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.680 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.700 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.720 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.740 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.760 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.780 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.800 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.820 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.840 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.860 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.880 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.900 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.920 101 128 0 0 100.0 100.0 100.0 0.0 0.0
0.940 97 128 0 4 98.3 96.0 100.0 0.0 3.0
0.960 97 128 0 4 98.3 96.0 100.0 0.0 3.0
0.980 90 128 0 11 95.2 89.1 100.0 0.0 7.9
1.000 0 128 0 101 55.9 0.0 100.0 . 44.1
Fit
Statistics for SCORE Data
Total Log Error
Data Set Frequency Likelihood Rate AIC AICC BIC
WORK.BAZA3 229 -1.0419 0 20.08389 20.90581 50.98739
Max-Rescaled Brier
Data Set SC R-Square R-Square AUC Score
WORK.BAZA3 50.98739 0.744175 0.996896 1 0.000157
Логит-регрессия
для двух групп CURE3Q в БД BAZA3 по п. 96 Программы 5
НЦ BIOSTATISTICA. Skype: leo_1911 E-mail: leo.biostat@gmail.com
Web-Site BIOMETRICA http:\\www.biometrica-tomsk.ru\ Tel: 8-989-830-83-10
_ _ _ _ S O
F F S 1 U T
_ A A E M S M N
P _ _ L L N S E S S
R P N P N S P S N P P
O O O E O E I E P S S S
b B S G S G T C E I E E
s _ _ _ _ _ _ _ C T N N
1 1.00000 47 128 0 54 0.46535 0.00000 1.00000 0.46535 1.46535 2.14894
2 1.00000 50 128 0 51 0.49505 0.00000 1.00000 0.49505 1.49505 2.02000
3 1.00000 54 128 0 47 0.53465 0.00000 1.00000 0.53465 1.53465 1.87037
4 1.00000 58 128 0 43 0.57426 0.00000 1.00000 0.57426 1.57426 1.74138
5 1.00000 61 128 0 40 0.60396 0.00000 1.00000 0.60396 1.60396 1.65574
6 1.00000 65 128 0 36 0.64356 0.00000 1.00000 0.64356 1.64356 1.55385
7 1.00000 69 128 0 32 0.68317 0.00000 1.00000 0.68317 1.68317 1.46377
8 0.99996 71 128 0 30 0.70297 0.00000 1.00000 0.70297 1.70297 1.42254
9 0.99978 75 128 0 26 0.74257 0.00000 1.00000 0.74257 1.74257 1.34667
10 0.99899 79 128 0 22 0.78218 0.00000 1.00000 0.78218 1.78218 1.27848
11 0.99631 83 128 0 18 0.82178 0.00000 1.00000 0.82178 1.82178 1.21687
12 0.99580 86 128 0 15 0.85149 0.00000 1.00000 0.85149 1.85149 1.17442
13 0.98841 90 128 0 11 0.89109 0.00000 1.00000 0.89109 1.89109 1.12222
14 0.98182 93 128 0 8 0.92079 0.00000 1.00000 0.92079 1.92079 1.08602
15 0.97694 97 128 0 4 0.96040 0.00000 1.00000 0.96040 1.96040 1.04124
16 0.93700 101 128 0 0 1.00000 0.00000 1.00000 1.00000 2.00000 1.00000
17 0.04182 101 124 4 0 1.00000 0.03125 0.96875 1.00000 1.96875 0.96875
18 0.03433 101 122 6 0 1.00000 0.04688 0.95313 1.00000 1.95313 0.95313
19 0.03419 101 119 9 0 1.00000 0.07031 0.92969 1.00000 1.92969 0.92969
20 0.02347 101 115 13 0 1.00000 0.10156 0.89844 1.00000 1.89844 0.89844
21 0.01603 101 112 16 0 1.00000 0.12500 0.87500 1.00000 1.87500 0.87500
22 0.00851 101 108 20 0 1.00000 0.15625 0.84375 1.00000 1.84375 0.84375
23 0.00587 101 104 24 0 1.00000 0.18750 0.81250 1.00000 1.81250 0.81250
24 0.00184 101 100 28 0 1.00000 0.21875 0.78125 1.00000 1.78125 0.78125
25 0.00010 101 96 32 0 1.00000 0.25000 0.75000 1.00000 1.75000 0.75000
26 0.00004 101 93 35 0 1.00000 0.27344 0.72656 1.00000 1.72656 0.72656
27 0.00000 101 0 128 0 1.00000 1.00000 0.00000 1.00000 1.00000 0.00000
Логит-регрессия
для двух групп CURE3Q в БД BAZA3 по п. 96 Программы 6
НЦ BIOSTATISTICA. Skype: leo_1911 E-mail: leo.biostat@gmail.com
Web-Site BIOMETRICA http:\\www.biometrica-tomsk.ru\ Tel: 8-989-830-83-10
_ _ _ _ S O
F F S 1 U T
_ A A E M S M N
P _ _ L L N S E S S
|
