different datatypes

-- Source: Ian Stirk, "Spring Clean Your Database Schema", SQLServerCentral – https://www.sqlservercentral.com/articles/spring-clean-your-database-schema

/*----------------------------------------------------------------------
 Purpose: Identify columns having different datatypes, for the same column name.
         Sorted by the prevalence of the mismatched column.
 ------------------------------------------------------------------------
 Revision History:
            06/01/2008  Ian_Stirk@yahoo.com Initial version.
 -----------------------------------------------------------------------*/
 -- Do not lock anything, and do not get held up by any locks.
 SET TRANSACTION ISOLATION LEVEL READ UNCOMMITTED
 -- Calculate prevalence of column name
 SELECT
       COLUMN_NAME
       ,[%] = CONVERT(DECIMAL(12,2),COUNT(COLUMN_NAME)* 100.0 / COUNT(*)OVER())
 INTO #Prevalence
 FROM INFORMATION_SCHEMA.COLUMNS
 GROUP BY COLUMN_NAME
 -- Do the columns differ on datatype across the schemas and tables?
 SELECT DISTINCT
         C1.COLUMN_NAME
       , C1.TABLE_SCHEMA
       , C1.TABLE_NAME
       , C1.DATA_TYPE
       , C1.CHARACTER_MAXIMUM_LENGTH
       , C1.NUMERIC_PRECISION
       , C1.NUMERIC_SCALE
       , [%]
 FROM INFORMATION_SCHEMA.COLUMNS C1
 INNER JOIN INFORMATION_SCHEMA.COLUMNS C2 ON C1.COLUMN_NAME = C2.COLUMN_NAME
 INNER JOIN #Prevalence p ON p.COLUMN_NAME = C1.COLUMN_NAME
 WHERE ((C1.DATA_TYPE != C2.DATA_TYPE)
       OR (C1.CHARACTER_MAXIMUM_LENGTH != C2.CHARACTER_MAXIMUM_LENGTH)
       OR (C1.NUMERIC_PRECISION != C2.NUMERIC_PRECISION)
       OR (C1.NUMERIC_SCALE != C2.NUMERIC_SCALE))
 ORDER BY [%] DESC, C1.COLUMN_NAME, C1.TABLE_SCHEMA, C1.TABLE_NAME
 -- Tidy up.
 DROP TABLE #Prevalence