Statistical inspection of food physical and chemical testing results (1)

1.

Two situations are often encountered in analysis and testing: the average value of the sample and the standard value of the sample are inconsistent; the average value of the two groups of measured data is inconsistent

1.

(1) Calculate the statistic t meter from the measured value

Where u——standard value of standard sample

x-the mean value of the sample

s——standard deviation of standard sample determination

n——The number of determinations of the standard sample

(2) According to the required confidence and degree of freedom (f=n-1), look up Table 1-17 to obtain the critical value t table of t

(3) Judgment:

t meter ≤ t table, indicating no significant difference from u

t meter> t table, which means there is a significant difference from u

Table 1-17 Critical/Value Table (Two-tailed)

Note: More critical t values ​​can be found in related books or manuals

[Example 1-3] In order to verify the accuracy of an analytical method, take 100 mg of the reference substance for 10 measurements

Solution: The calculated average and standard deviation: =99.

Check table 1-17: f=n-1=9, when the significance level α=0.

Judgment: 2.

[Example 1-4] Calibrate the instrument with a standard solution with a standard value of 1.

Solution: Known from the question=1.

Check table 1-17: n=10, f=n-1=9, and the confidence level is 95%, that is, when the significance level a=0.

Judgment: t meter = 2.

2.

It is used for comparison between two sets of data measured by two methods (new method and standard method), between two groups of data measured by two analysts or two laboratories
.

Inspection steps:

(1) According to the mean 1, 2, and standard deviations s ₁ and s _{2 of the} two sets of data , find the combined standard deviation s sum
.

(2) Calculate the statistic t meter
.

(3) Check critical value: given a value, degree of freedom f=n ₁ +n ₂ -2, check table 1-17 to get t table
.

(4) Judgment:

t meter ≤ t table, there is no significant difference between the two means

t meter> t table, the two mean values ​​are significantly different

Related links: The original record of food physical and chemical testing data and the filling of the testing report