AOAC Methods in Codex STAN 234 (Preliminary Methods Review)
856
BARBANO ET AL.: J. ASSOC. OFF. ANAL. CH!'.M. (VOL. 73, NO. 6, 1990)
Table 2. AOAC 11tallstlcal parameters by sample materlal for percent protein determined by KJeldahl m.thod In fltsl lrtal or collaborallve study (digested for 1 h aNer cl.arlng) Material
7
4
5
9
2
3
6
8
All
All
Outlier
Outlier
Mean•
date
data
StatlltJc
••eluded
excll.ldlJd
II
9
9
10
10
10
10
10
9
10
No. of labs
9
No. of accepted values
18
18
20
20
18
18
20
16
20
20
20
3.367 0.021 0.810 0.024 0.706 0.058 0.087
3.396 0.011 0.336 0.015 0.441
3.086 0.007 0.237 0.016 0.529 0.021 0,048
3.610
3.082
3.091 0.025 0.817 0.023 0.737 0.071 0 .084
3.140 0.007 0.225 0.012 0.387
3.259 0.013 0,385 0.019 0.587
3.379 0.046 1.359 0.054 1.588 0.130 0.152
3.266 0.018 0.540 0.020 0.611 0.050 0.058
3.296 0.015 0.446 0.020 0.611 0.042 0.057
3.446 0.013 0,384 0.026 0.739
Mean,%
s,
0.018 0.045
1.445 0.046 1,479 0.126 0.129
0.490 0.027 0.760 0.050 0.078
RSD,, %
SR
RSDA, % r Value R Value
0.020 0 ,036
0.035 0.032
0.054
0.042
0.034
0.072
• Means are baaed on dal.a from sample rne.terlals 5 and 8 with outllers excluded.
regression.~ of percent protein vs each AOAC statistical pa· rametcr for each type of equipment found no significant influence of protein concentration on any of the traditional equipment AOAC statistical parameters but there was a significant (P < 0.05) influence of protein concentration on SR and R values for the block digcstor equipment. One factor that may cause block digestors in different laboratories to display different recovery of nitrogen on high protein samples at the end of a fixed (I h) time of boiling after clearing may be laboratory-to-laboratory variation in line voltage. Differences in line voltage can resuit in different block temperature recovery times, as indicated by the manu– facturer {P. Calvi, Tecator, Inc., personal communication). The provisional JDF procedure requires that boiling be con– tinued for 1.5 h after clearing (the method for our first
collaborative study trial was I h after clearing). If digestion heating conditions differed between laboratories usmg block digestors, digestion could be more complete in some labora– tories than in other laboratories. We hypothesized that sam– ple digestion may not have been complete in all laboratories using block digestors. This could explain the increase in the laburatory-to-Jaboratory varjability of test results with in– creasing protein content in the first trial. Thus, a second trial in the collaborative study was conducted to determine if a boiling time of I .5 h after clearing would eliminate the corre- lation between protein concentration and reproducibility sta- tistics. Relative normality of the titrant used in each laboratory I d d . .f . . Ii was eva uate to etermine 1 errors in htrant norma ty contributed to differences in mean test results from Jaborato-
Tabte 3. Linear regr...ion ot ,,, RSD,, •R, ASDR, r value, and R value va percent proteJn (all ragrealone comprised 9 test t values with 7 d99,.., of freedom) tor flm trlal In collaboratlve 1tudy
Corr. coetf., %
.,
Mean
Eq\llpment
S1atlstic
Slope
Intercept
~Value
b
0.015 0.445 0.020 0.811 0.042 0.057 3.296 0.015 0.449 0.018 0.545 0.042 0.051 S.292 0.013 0.388 0.022 0,663 O.G36 0.062 3.302
0.0051 0.0158 0.0176 0.3347 0.0145 0.0-499 0.0105 0.1679 0.0346 0.8640 0.0294 0.0979
-0.0021
All
2.19 0.02
0.396 0 .039 2.034 1.282 0.396 2.034 0.732 0.383 2.784· 2.300 0.732 2.784·
RSD,, %
0.3927
-0.0379 -0.4922 -0.0061 -0.1072
37.15 18.52
8A
RSDA, %
r value R value
2.18
37.15
% protein
a,.
-0.0190 -0.1040 -0.0950 -2.2990 -0.0540 -0.2710
7.10 2.04
Block
RSD,, %
52.54 43.03
8R
RSOn, %
r value R value
7.10
N N
52.54
% protein
., R Sf!
M
- 0.0104 -0.4223
0.0473 1.7819 0.0097 0.9872 0.1 338 0.0274
4.72 7.60 4.12 3.03 4.72 4.12
-0.589 -0.759
Tredlllonal
Sr
ASD,, %
0.0037
0.549
Bil
-0.0IIBO -0.0296
-0.468 -0.589
RSOR, %
AS r V R i;
r value R value
0.0104
0 ,549
% protein
Codex Trial Method Review • Significant at a • o.os.
Made with FlippingBook - Online catalogs