Una forma de hacerlo es usar ROW_NUMBER
función con partición para detectar cuando el valor del Department
cambios de columna.
Datos de muestra
DECLARE @T TABLE (ID int, Department nvarchar(100), dt date);
INSERT INTO @T (ID, Department, dt) VALUES
(1, 'English', 'Feb 3 2017'),
(1, 'English', 'Feb 4 2017'),
(1, 'Science', 'Mar 1 2017'),
(1, 'Science', 'Apr 2 2017'),
(1, 'Maths ', 'Apr 7 2017'),
(2, 'Maths ', 'Feb 1 2017'),
(2, 'Maths ', 'Apr 7 2017'),
(3, 'Maths ', 'Apr 3 2017'),
(4, 'Science', 'Feb 1 2017'),
(4, 'Maths ', 'Apr 7 2017');
Consulta
WITH
CTE
AS
(
SELECT
ID
,Department
,dt
,ROW_NUMBER() OVER (PARTITION BY ID, Department ORDER BY dt DESC) AS rnPart
,ROW_NUMBER() OVER (PARTITION BY ID ORDER BY dt DESC) AS rnID
FROM @T
)
SELECT
ID
,Department
,dt
FROM CTE
WHERE
rnPart = 1
AND rnID <> 1
ORDER BY
ID
,dt
;
Resultado
+----+------------+------------+
| ID | Department | dt |
+----+------------+------------+
| 1 | English | 2017-02-04 |
| 1 | Science | 2017-04-02 |
| 4 | Science | 2017-02-01 |
+----+------------+------------+