To conduct a systematic analysis of the safety risks associated with edible agricultural products by applying statistical methods.

Based on the supervision and sampling inspection data of edible agricultural products in 18 areas of Beijing in 2025 released by the State Administration for Market Regulation, 7 common kinds of edible agricultural products including hot pepper, celery, scallion, ginger, pork, eggs and marine fish were selected as the study objects for safety risk analysis. The binomial test with 2 benchmarks set at p₀=0.99, p₀=0.95 was adopted to classify risk levels, and special analysis was conducted focusing on the residue profiles of prohibited and restricted pesticides, veterinary drugs and heavy metals. Meanwhile, the Chi-square test was used to conduct statistical analysis on the risk differences among different varieties and the residue level differences in different regions.

The overall nonconformance rate of the selected 7 kinds of food was 0.15% (3/1942), and only scallion and egg were non-compliant, with pesticide thiamethoxam and veterinary drug florfenicol (the sum of florfenicol and florfenicol amine combined) respectively. The binomial test identified 2 kinds of potential high-risk varieties (hot pepper and celery,P<0.05 at p₀=0.99) and 5 kinds of potential medium-risk varieties (scallion, ginger, pork, egg and marine fish, P<0.05 at p₀=0.95), with no low-risk varieties found. The Chi-square test indicated no significant difference in risks among the categories (χ²=9.6135, df=6, P=0.1419>0.05). The detection rate of thiamethoxam in scallions was 12.8% (34/266) and the excess rate was 0.75% (2/266), and the excessive samples were concentrated in Fangshan District; only 1 sample of egg was excessive in florfenicol (0.44%, 1/226), which was also from Fangshan District. Regional difference analysis showed that the detection rate of thiamethoxam in scallion in Fangshan District (34.4%) was significantly higher than that in Fengtai District (3.8%), and statistically significant regional differences existed in the detection rate of florfenicol in eggs (P<0.05).

The safety risk analysis method for edible agricultural products based on the binomial test and Chi-square test constructed in this study can effectively make up for the defect that traditional risk assessment overrelies on the actual nonconformance rate, and provide a reliable statistical tool for the risk classification and precise supervision of the quality and safety of edible agricultural products. The study results can provide data support for market supervision departments to optimize sampling inspection schemes. It is suggested that the sampling inspection frequency should be increased for potentially high-risk varieties. Meanwhile special rectification of pesticide residues should be carried out on scallion products, the source control in key areas and the supervision in circulation links should be strengthened, so as to effectively ensure the quality and safety of edible agricultural products and protect the legitimate rights and interests of consumers.