Table 8 ANOVA for COVID-19 cases per day in New York, Kasaragod, Srinagar and Mumbai.

From: Exploring dependence of COVID-19 on environmental factors and spread prediction in India

(i) Source

DF

Seq SS

Adj SS

Adj MS

F

P

DF

Seq SS

Adj SS

Adj MS

F

P

 

a) New York

c) Srinagar

Regression

5

21.0059

21.0059

4.2012

29.23

<0.001

5

6.3503

6.35034

1.27007

24.84

<0.001

Linear

2

18.2906

18.2801

9.14

63.59

<0.001

2

5.0432

3.2961

1.64805

32.24

<0.001

X1

1

11.9274

13.881

13.881

96.58

<0.001

1

5.0432

0.36702

0.36702

7.18

0.009

X2

1

6.3632

2.4437

2.4437

17

<0.001

1

0

0.00688

0.00688

0.13

0.714

Square

2

1.1471

0.3321

0.1661

1.16

0.318

2

1.3003

0.02696

0.01348

0.26

0.769

X1*X1

1

0.2846

0.0841

0.0841

0.58

0.446

1

1.147

0.02307

0.02307

0.45

0.503

X2*X2

1

0.8625

0.2592

0.2592

1.8

0.182

1

0.1533

0.00395

0.00395

0.08

0.782

Interaction

1

1.5683

1.5683

1.5683

10.91

0.001

1

0.0068

0.00678

0.00678

0.13

0.717

X1*X2

1

1.5683

1.5683

1.5683

10.91

0.001

1

0.0068

0.00678

0.00678

0.13

0.717

Residual Error

126

18.11

18.11

0.1437

  

104

5.3168

5.31679

0.05112

  

Total

131

39.116

    

109

11.6671

    
 

S = 0.379118 PRESS = 19.8260 R-Sq = 53.70% R-Sq(adj) = 51.86%

S = 0.226104 PRESS = 5.88259 R-Sq = 54.43% R-Sq(adj) = 52.24%

 

b) Kasaragod

d) Mumbai

Regression

5

1.52228

1.52228

0.304455

2.97

0.018

5

6.4983

6.4983

1.2997

7.63

<0.001

Linear

2

0.96293

1.12665

0.563323

5.5

0.006

2

2.0294

3.8953

1.9477

11.44

<0.001

X1

1

0.91382

0.8544

0.854401

8.35

0.005

1

0.0029

3.6813

3.6813

21.62

<0.001

X2

1

0.04911

0.00219

0.002187

0.02

0.884

1

2.0266

3.242

3.242

19.04

<0.001

Square

2

0.29411

0.55667

0.278335

2.72

0.073

2

3.8473

2.9026

1.4513

8.52

<0.001

X1*X1

1

0.28967

0.5446

0.544596

5.32

0.024

1

3.816

2.4865

2.4865

14.6

<0.001

X2*X2

1

0.00444

0.12586

0.125857

1.23

0.272

1

0.0312

0.2367

0.2367

1.39

0.241

Interaction

1

0.26524

0.26524

0.265236

2.59

0.112

1

0.6217

0.6217

0.6217

3.65

0.059

X1*X2

1

0.26524

0.26524

0.265236

2.59

0.112

1

0.6217

0.6217

0.6217

3.65

0.059

Residual error

65

6.65327

6.65327

0.102358

  

115

19.5837

19.5837

0.1703

  

Total

70

8.17554

    

120

26.082

    
 

S = 0.197500 PRESS = 6.53204 R-Sq = 47.81% R-Sq(adj) = 45.44%

S = 0.412665 PRESS = 21.6522 R-Sq = 24.91% R-Sq(adj) = 21.65%

 

New York

Kasaragod

(ii) Terms

Coef

SE Coef

T

P

Coef

SE Coef

T

P

Constant

−0.3152

0.07739

−4.073

<0.001

−0.8546

0.03959

−21.586

<0.001

X1

−0.5623

0.05722

−9.827

<0.001

−0.3645

0.18301

−1.992

0.049

X2

0.3747

0.09088

4.123

<0.001

−0.2049

0.16127

−1.27

0.207

X1*X1

−0.1044

0.13654

−0.765

0.446

1.5416

0.25872

5.959

<0.001

X2*X2

−0.2214

0.16486

−1.343

0.182

1.5716

0.22252

7.062

<0.001

X1*X2

−0.44

0.13321

−3.303

0.001

3.0761

0.3548

8.67

<0.001

 

Srinagar

Mumbai

 

Coef

SE Coef

T

P

Coef

SE Coef

T

P

Constant

−0.84597

0.04004

−21.13

<0.001

−0.4926

0.08659

−5.689

<0.001

X1

0.3265

0.12186

2.679

0.009

−0.9608

0.20665

−4.649

<0.001

X2

−0.05388

0.14684

−0.367

0.714

0.7448

0.1707

4.363

<0.001

X1*X1

0.14775

0.21996

0.672

0.503

1.3213

0.34578

3.821

<0.001

X2*X2

0.10358

0.37282

0.278

0.782

0.3022

0.25631

1.179

0.241

X1*X2

−0.18598

0.51086

−0.364

0.717

−1.0198

0.53373

−1.911

0.059

  1. The Analysis of Variance (ANOVA) depicts the significance of the linear, squared and interaction terms of Temperature and Relative Humidity on the response, i.e., COVID-19 cases per day. Here, DF represents the degree of freedom, Seq SS represents sequential sum of squares, Adj SS is Adjusted sum of squares and Adj. MS represents Adjusted Mean Square and all these terms, help determine the variability capture by each model terms in the RSM quadratic model. The RSM model fitness is captured by R-sq value and the Adj. R-sq values within 10% of the R-sq values confirm the model adequacy. High value of Tstatistics > Tcriticial and p < 0.05 at confidence interval of 95% denotes significant impact of the model term on the response.
Back to article page