面对政府统计工作改革和行业决策管理的双重要求，针对公路货物运输量统计数据受外部 因素干扰，存在较严重准确性的问题，从控制省级上报统计数据质量角度入手，在研究不同评估方法对公路货运量的适用性基础上，提出了一种基于相关性和正态分布的异常值检验方法。从行 业数据内部协调性和外部数据匹配性的角度，选取了高度相关的判别指标作为评估依据，构造了判别指标与货运量的发展速率比服从正态分布的检验统计量。在给定的显著性水平下计算该统计 量并进行异常值诊断，检验各省货运量增长速度的截面数据是否偏离正态分布成为异常值，以此建立公路运输量统计评估模型，为开展公路货物运输量数据质量评估工作提供指导。结果表明，在实际工作中，该方法可以有效诊断出数据异常省份并指导其进行数据核实和修正，有助于建立长期质量评估机制，从而反映公路运输业真实发展情况，为行业提供更有效、更可靠的统计信息 服务。

Facing the double requirements of government statistical work reform and industry decision management, in regard with the problem of poor statistical data quality of highway freight volume by external interfering elements, starting from the point of controlling the quality of the statistical data at the provincial level, a test method of abnormal data based on correlation and normal distribution was pro⁃ posed on the basis of analyzing applicability of different assessment methods to highway freight volume. By selecting highly correlated distinguishing indicators as basis for assessment, from the perspective of internal coordination of industry data and the matching of external data, a test statistic was established based on the implicit assumption that development rate ratio between distinguishing indicators and high⁃ way freight volume should obey normal distribution. Then the statistic was calculated at a given level of significance and the abnormal data was diagnosed, aiming for check whether the cross-sectional data of the highway freight volume of each province deviated from the normal distribution as abnormal data. A statistical assessment model of highway freight volume was established to guide the data quality assess⁃ment of highway freight volume. The results demonstrate that the proposed method could effectively diag⁃ nose provinces with abnormal data and guide them to verify and correct data in practical work, and contribute to establish long-term quality assessment mechanism, which makes sense for reflecting the real development of highway transportation industry and providing more effective and reliable statistical information services for the industry.