Statistics Meeting Book (June 20, 2018)

AOAC O FFICIAL M ETHODS OF A NALYSIS (2012)

M ICROBIOLOGY G UIDELINES Appendix J, p. 17

ANNEX E Example of Graph of POD Values from Qualitative Method Single Laboratory Data

Summary Data Table for Qualitative Method Single Laboratory Studies Table D1. Comparative results for the detection of Listeria innocua in raw shrimp Concentration Candidate presumptive (CP) Candidate confirmed (CC) Candidate method (C) Reference method (R) C vs R CP vs CC Statistic MPN/25 g N x POD(CP) N x POD(CC) N x POD(C) N x POD(R) dPOD(C,R) dPOD(CP,CC) Estimate 0.00 20 0 0.00 20 0 0.00 20 0 0.00 20 0 0.00 0.00 0.00 LCL 0.00 0.00 0.00 0.00 0.00 –0.16 –0.16 UCL 1.50 0.16 0.16 0.16 0.16 0.16 0.16 Estimate 0.80 20 12 0.60 20 10 0.50 20 10 0.50 20 11 0.55 –0.05 0.10 LCL 0.43 0.39 0.30 0.30 0.34 –0.33 –0.19 UCL 1.39 0.78 0.70 0.70 0.74 0.24 0.37 Estimate 3.00 20 20 1.00 20 20 1.00 20 20 1.00 20 19 0.95 0.05 0.00 LCL 1.58 0.84 0.84 0.84 0.76 –0.12 –0.16 UCL 5.68 1.00 1.00 1.00 1.00 0.24 0.16 Estimate 17.00 20 20 1.00 20 20 1.00 20 20 1.00 20 20 1.00 0.00 0.00 LCL 0.27 0.84 0.84 0.84 0.84 –0.16 –0.16 UCL 1060 1.00 1.00 1.00 1.00 0.16 0.16

Notes : ( 1 ) The concentration plotted should be MPN/test portion. ( 2 ) Confidence intervals may also be plotted. ( 3 ) Collaborative data should be plotted analogously. ANNEX F Calculation of LPOD and dLPOD Values from Qualitative Method Collaborative Study Data For a multilaboratory trial where L = number of laboratories, R = replicates per laboratory, N = LR = total replicates, LPOD estimate is given by where x is the number of positive results. Method for estimating LPOD 95% confidence intervals: Step 1. —Enter data into AOAC spreadsheet with 1 for positive response and 0 for negative response. Record the mean LPOD, s(R), and s(r). Step 2 .—Calculate s(L), standard deviation due to laboratory effect as: ( ) ( ) s L s R s r Step 3 .—Calculate s(POD) as the standard deviation of the individual laboratory POD estimates. 2 ( ) ( ) 1 i POD LPOD s POD L ¦ Step 4 .—Calculate degrees of freedom, df for s(POD) as follows: 2 2 ( )

ANNEX D

2 2

ª « « ¬

º » » ¼

s r

2

s L

( )

L N

df

2 2

2 2 º « » « ¼ ¬ ª

º » » ¼

s r

ª « ¬

s L

( )

N

L

L

N L

1

Step 5 .—Calculate 95% confidence limits on LPOD: If 0.15 ≤ LPOD ≤ 0.85:

© 2012 AOAC INTERNATIONAL