Data Planet LibGuides Blog

Standardization and Data

In analyzing statistical results, it is important to make sure that "apples are being compared to apples"—meaning that an apple is defined clearly in both data collection and data reporting to mean the same thing. When possible, variables should be defined in a way that is consistent with commonly used definitions or classification systems. Some of these systems apply cross-nationally, particularly in the area of international trade, while others are set at the national level. Two examples from North America:

1. In the United States, the Office of Management and Budget (OMB) is charged by statute with coordinating the federal statistical system, and one of its key responsibilities is to establish standards and guidelines for data collection and dissemination. One such standard developed under OMB auspices is the North American Industrial Classification System (NAICS) by which the industry of a business establishment is classified based on its primary business activity. The NAICS system  here was created and adopted in 1997 to allow for comparability across the business economies in the United States, Canada, and Mexico. NAICS is used by the statistical agencies of the US, Canada, and Mexico, although some national-level adaptations have been incorporated: Statistics Canada has created five cannabis industries that are unique to NAICS Canada 2017 Version 3.0.

2. US government agencies also follow the OMB 1997 Revision to the Standards for the Classification of Federal Data on Race and Ethnicity (Directive No. 15) . The directive describes the classifications as promoting uniformity and comparability for data on race and ethnicity. This particular standard is specific to the US. Statistics Canada follows a 2015 standard that relies on the concept of “visible minorities,” defined in the Employment Equity Act (S.C. 1995, c. 44) .

Use of such standards helps ensure comparability of results across surveys and time. When evaluating a dataset, be sure to consider whether the variables have been standardized, so that you can be confident in the comparability of the statistics you use in your analysis!