Table 4 Comparison of statistical parameters of MLR models with HA index as an outcome and standardized predictors

From: A simple generalized mass transfer model for evaluation of the environmental performance of archival boxes

 

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

Boxes included in the model

Empty (unstacked) Boxes

Empty&Filled (unstacked) Boxes

Empty&Stacked Boxes

Empty&Filled& Stacked Boxes

Empty&Filled& Stacked Boxes

Empty&Filled& Stacked Boxes

Number of boxes

43

48

48

53

53

53

(Intercept)

69.7963

72.7933

72.6777

75.1200

75.1200

75.1202

Std. Error

0.9565

0.9387

0.8781

2.471

2.196

0.8593

Pr(>|t|)

<2E-16

<2E-16

<2E-16

<2E-16

<2E-16

<2E-16

(AER/MSI)stand.

−15.6872

−15.1213

−15.3642

−17.3520

 

−14.7719

Std. Error

0.9713

0.9551

0.8950

2.495

 

0.8778

Pr(>|t|)

<2E-16

<2E-16

<2E-16

6.41E-9

 

<2E-16

(WVTR/MSD)stand.

−16.8443

−17.1988

−16.7644

 

−19.177

−16.9257

Std. Error

0.9713

0.9551

0.8950

 

2.217

0.8778

Pr(>|t|)

<2E-16

<2E-16

<2E-16

 

1.43E-11

<2E-16

Residual Std. Error

6.272

6.503

6.083

17.99

15.99

6.256

Multiple R-squared

0.9387

0.9353

0.9428

0.4868

0.5946

0.9392

p-value

<2E-16

<2E-16

<2E-16

6.415E-9

1.429E-11

<2E-16

RMSE

6.0496

6.2969

5.8902

17.6483

15.6855

6.0760

RMSELOOCV(a)

7.4606

7.5156

7.1708

18.2351

16.2365

7.2047

RMSErepeated k-fold CV(b)

7.6143

7.6142

7.3146

18.2972

16.3497

7.2912

  1. aRMSE – Leave-One-Out Cross-Validation.
  2. bRMSE – 500 repetitions, data split into 5 groups (folds).