Table 6 Performance re-evaluation: linear regression on results from bioanalytical screening (BEQ) of 25 pig meat (pork) samples for PCDD/Fs and DL-PCBs, and from GC/HRMS analysis (TEQ), followed by calculation of the respective cut-off concentrations, and of α- and ß-errors (false-noncompliant and false-compliant rates) [18, 22]

Number of samples

N

19

19

19

19

Number of analyses per sample^a

m

1 (6 × 2)

1 (6 × 2)

1 (6 × 2)

1 (6 × 2)

Number of analyses (total)

n

25

25

25

25

Calibrated range

pg TEQ/g fat

R

0.07–2.86

0.07–2.86

0.23–3.59

0.30–5.67

Slope, calibration line

b

0.904

0.904

0.722

0.891

Standard error of the slope

s_b

0.050

0.050

0.032

0.029

y-intercept, calibration line

pg BEQ/g fat

a

0.092

0.092

0.134

− 0.060

Standard error of the y-intercept

pg BEQ/g fat

s_a

0.075

0.075

0.063

0.093

Coefficient of correlation

r

0.9665

0.9665

0.9784

0.9884

Critical value for r (df = n − 2, p = 0.05, two-sided)

crit_r

0.3961

0.3961

0.3961

0.3961

Square coefficient of correlation

r²

0.9342

0.9342

0.9573

0.9770

GC/HRMS decision limit^b

pg TEQ/g fat

DL

1.2

0.90

0.60

1.50

BEQ concentration at DL

pg BEQ/g fat

y_DL

1.2

0.91

0.57

1.28

Residual standard deviation^c

pg BEQ/g fat

s_yx

0.2326

0.2326

0.1504

0.2155

Relative s_yx at y_DL (precision^c)

%

s_yx,rel

19.8

25.7

26.5

16.9

Cut-off concentration

pg BEQ/g fat

y_Cut-off

0.70

0.42

0.24

0.82

α-Error (false-noncompliant rate)^d,e

%

α

17 / 0

38 / 24

0 / 0

10 / 8

ß-Error (false-compliant rate)^d

%

ß

0

0

0

0

Maximum level

pg TEQ/g fat

ML

1.0

–

–

1.25

Action level

pg TEQ/g fat

AL

–

0.75

0.50

–