HardSQL challenge #251WINDOW FUNCTIONPARTITIONRANKING

Find the Second Highest Salary in Each Department

Return the employees who earn the second-highest salary of their department.

The challenge runs on the SQLabHub practice database (SQLite) and uses the employees table. You can solve it in the browser, with nothing to install.

Solve This SQL Challenge

The task

Find employees with second highest salary in each department.

A correct answer returns 10 rows with 13 columns: id, first_name, last_name, email, department, department_id, title, salary, hire_date, manager_id, city, status, rnk.

Tables used in this challenge

`employees` 120 rows

id INTEGER (PK) · first_name TEXT · last_name TEXT · email TEXT · department TEXT · department_id INTEGER · title TEXT · salary REAL · hire_date TEXT · manager_id INTEGER · city TEXT · status TEXT

First 3 rows of the employees table
id	first_name	last_name	email	department	department_id	title	salary	hire_date	manager_id	city	status
1	William	Lewis	wlewis18@corp.io	Customer Success	7	Senior CS Manager	61774	2022-07-24	NULL	San Diego	active
2	John	Miller	jmiller51@corp.io	Legal	8	Paralegal	191694	2021-08-09	NULL	San Diego	active
3	Daniel	Rivera	drivera50@corp.io	Engineering	1	VP Engineering	135067	2016-06-06	NULL	Houston	active

The first rows of each table are shown. See how all the tables relate on the database schema page.

Expected output

Expected output of the challenge
id	first_name	last_name	email	department	department_id	title	salary	hire_date	manager_id	city	status	rnk
60	Maria	Garcia	mgarcia35@corp.io	Customer Success	7	CS Manager	191120	2016-11-09	9	Washington DC	active	2
73	Joshua	Hall	jhall66@corp.io	Design	10	UI Designer	191657	2015-10-19	4	San Francisco	active	2
49	Dorothy	Martinez	dmartinez19@corp.io	Engineering	1	Engineering Manager	195690	2021-04-21	5	Philadelphia	active	2
65	Linda	Martinez	lmartinez23@corp.io	Finance	5	Financial Analyst	194604	2015-07-14	3	Philadelphia	active	2
102	Michelle	Rodriguez	mrodriguez67@corp.io	HR	4	HR Manager	172167	2020-04-22	2	Jacksonville	active	2
2	John	Miller	jmiller51@corp.io	Legal	8	Paralegal	191694	2021-08-09	NULL	San Diego	active	2
26	Matthew	Wilson	mwilson27@corp.io	Marketing	2	CMO	196530	2015-10-03	3	Columbus	active	2
82	Joshua	Nguyen	jnguyen7@corp.io	Operations	6	COO	182684	2023-05-06	4	Austin	active	2
118	Michael	Young	myoung35@corp.io	Product	9	CPO	166288	2015-02-16	3	Los Angeles	active	2
57	Paul	Thompson	pthompson52@corp.io	Sales	3	Sales Manager	168980	2022-01-05	7	Fort Worth	active	2

The complete result: 10 rows. This is the output of the model solution on the practice database. The task does not fix a sort order, so your rows may come back in a different order.

What you practise

Filtering rows with a WHERE condition
Querying the result of another query with a subquery in FROM
Ranking rows with RANK() or DENSE_RANK() and handling ties
Restarting a window calculation for each group with PARTITION BY

Hint and approach

Try the challenge first. Open these only when you need them.

Show the hint

DENSE_RANK

Show the step-by-step approach

How the model solution is built, in the order the database evaluates it:

Build an inner query first (a derived table):
1. Read from the employees table.
2. Rank the rows with DENSE_RANK(), ordered by salary descending, restarting for each department. Tied rows share a rank and no rank number is skipped.
Filter in WHERE: keep the rows where rnk equals 2. The filter sits in the outer query because a window function cannot be tested in the WHERE clause of the query that computes it.

Model solution

Checking your session…

What to watch for

DENSE_RANK() = 2 picks the second salary level even when several people share the top one. A department with a single salary level returns nobody.

RANK() and DENSE_RANK(). RANK() skips numbers after a tie (1, 1, 3) and DENSE_RANK() does not (1, 1, 2). Filtering on rank <= N returns more than N rows when there are ties.
PARTITION BY. PARTITION BY is not GROUP BY: it keeps every row and only decides which rows a window function looks at.
Derived table. Columns computed in the inner query are only visible outside under their alias, so every calculated column there needs a name.