Skip to main content

Table 5 Alternative estimator results for log-normal, gamma and weibull distributions for n=500

From: Statistical models for the analysis of skewed healthcare cost data: a simulation study

Data

Estimator

MPE

MPAE

MSE(β)

95% CI

AIC

Prob. H.Lsignif

Lower

upper

Log normal σ2=0.5

OLS for Ln(y)

-0.00617

0.55823

0.01166

0.991

1.011

1075.552

0.0438

Gamma

-0.000002

0.55662

0.01079

0.989

1.011

830.756

0.0405

Weibull

-0.11093

0.59335

0.01155

0.987

1.011

870.429

0.0538

Cox

-1.31086

3.10119

5.36566

-1.326

-1.296

5157.713

0.0490

Log normal σ2=1

OLS for Ln(y)

-0.00625

0.76743

0.02331

0.987

1.015

1422.125

0.0444

Gamma

-0.00002

0.76539

0.02041

0.981

1.017

953.996

0.0380

Weibull

-0.00211

0.76577

0.02309

0.982

1.016

958.951

0.0382

Cox

-0.92086

3.01376

3.71427

-0.935

-0.907

5189.630

0.0543

Log normal σ2=1.5

OLS for Ln(y)

-0.00646

0.92875

0.03497

0.985

1.019

1624.858

0.0406

Gamma

-0.00004

0.92652

0.02935

0.974

1.022

945.716

0.0338

Weibull

0.12644

0.87192

0.03464

0.978

1.020

919.644

0.0351

Cox

-0.74999

2.98739

3.08671

-0.764

-0.736

5200.723

0.0474

Log normal σ2=2

OLS for Ln(y)

-0.00665

1.05164

0.04662

0.983

1.021

1768.699

0.0407

Gamma

-0.00006

1.04944

0.03788

0.966

1.028

867.320

0.0316

Weibull

0.25187

0.93223

0.04619

0.974

1.024

813.451

0.0371

Cox

-0.64857

2.97510

2.74186

-0.663

-0.635

5206.363

0.0500

Gamma α=0.5

OLS for Ln(y)

-0.01173

0.97145

0.170

0.966

1.026

2218.380

0.0814

Gamma

-0.00010

0.96635

0.069

0.981

1.019

745.079

0.0395

Weibull

0.26621

0.88808

0.082

0.979

1.018

756.009

0.0613

Cox

-0.69386

3.04111

2.896

-0.999

-0.388

5204.358

0.050

Gamma α =1

OLS for Ln(y)

-0.00739

0.73625

0.056

0.984

1.018

1669.842

0.0582

Gamma

-0.00001

0.73405

0.034

0.987

1.014

960.724

0.0431

Weibull

-0.00095

0.73423

0.034

0.987

1.014

960.723

0.0438

Cox

-1.00444

3.10634

4.035

-1.019

-0.990

5184.427

0.0468

Gamma α =2

OLS for Ln(y)

-0.00643

0.54150

0.022

0.999

1.013

1202.164

0.0452

Gamma

-0.00002

0.54021

0.017

0.992

1.011

844.867

0.0403

Weibull

-0.10982

0.56708

0.016

0.992

1.011

851.287

0.0546

Cox

-1.42736

3.23880

5.909

-1.442

-1.413

5148.590

0.0461

Gamma α =4

OLS for Ln(y)

-0.00606

0.39091

0.010

0.993

1.007

792.221

0.0443

Gamma

0.000004

0.39006

0.008

0.993

1.007

598.026

0.0416

Weibull

-0.13200

0.42060

0.008

0.993

1.007

617.434

0.1017

Cox

-2.01502

3.48489

9.092

-2.031

-1.999

5086.403

0.0486

Wiebull α=0.5

OLS for Ln(y)

-0.01379

1.18150

0.15321

0.962

1.032

2362.321

0.0606

Gamma

-0.00012

1.17416

0.06475

0.965

1.025

411.304

0.0338

Weibull

0.49762

0.97207

0.09307

0.969

1.025

384.861

0.0693

Cox

-0.49022

2.99166

2.25145

-0.563

-0.421

5213.082

0.0495

Wiebull α =1

OLS for Ln(y)

-0.00741

0.73714

0.03830

0.980

1.016

1669.173

0.0530

Gamma

-0.00002

0.73494

0.02327

0.984

1.012

961.400

0.0421

Weibull

-0.00082

0.73506

0.02326

0.984

1.012

961.376

0.0418

Cox

-0.99154

3.11589

4.00036

-1.066

-0.922

5184.367

0.0473

Wiebull α =5

OLS for Ln(y)

-0.00605

0.18346

0.00153

0.996

1.004

59.7355

0.0453

Gamma

-0.000003

0.18362

0.00138

0.997

1.003

-51.535

0.0447

Weibull

-0.08896

0.19356

0.00093

0.997

1.003

-101.476

0.2244

Cox

-5.00813

6.36391

36.15827

-5.029

-4.987

4737.774

0.0530