The CO may begin the desk audit analysis of compensation once the contractor’s compensation data submission is complete and acceptable. What constitutes an acceptable submission is discussed in Section 1F05. Analysis should begin with an examination of the descriptive statistics related to the data. Descriptive statistics are fact-based statistics such as mean, median, modes, or other distributional measures. The CO may use preliminary tools, tests, or broad indicators of potential compensation issues at the desk audit to identify pay practices or issues needing further review. In most cases, it is insightful to use nonsystemic comparative analysis to compare the treatment of similarly situated individuals or small groups of employees. In familiarizing themselves with the data, the CO should ascertain whether groupings, ratings, rankings, labels, and other employer-assigned attributes are differentially distributed by gender, race, or national origin. As with the systemic analysis of compensation, the CO must ensure that similarly situated employees are being compared and must assess whether any explanatory factors the contractor offers are implemented fairly, consistently applied, and relevant to the contractor’s compensation system. This data exploration stage also includes an assessment of the pay analysis groups for reasonableness and statistical sufficiency. Considering these questions in consultation with supervisors, RSOL and Branch of Expert Services (BES) will allow the CO to determine whether initial indicators warrant further systemic analysis and on-site investigation.