Data integrity means keeping data accurate, consistent, complete, and valid while preparing it for analysis. In a basketball salary dataset, integrity checks begin by retaining only needed variables, converting categories to factors, and inspecting summaries for blanks and uneven coverage. section 12.4
A practical integrity workflow includes:
- Select salary, season_end, and team instead of carrying unnecessary variables. section 12.4
- Convert season_end and team to factors, then use summary() to inspect distributions and group counts. section 12.4
- Investigate four blank team values and recognize that early seasons have unusually few salary records. section 12.4
- Exclude seasons from 1985 through 1990 before ranking salaries within each season and team. section 12.4
Checking suspicious records
A SQL query finds four records with blank team values, all from 1985. Comparing Boston Celtics records from 1987 and Los Angeles Lakers records from 1990 confirms that some early team-season groups contain only three or four observations, limiting reliable comparison. section 12.4
