LMS Competency Tests

Competency tests

New in 3.0.0 · LMS → Tests · permission: Competency tests (sensitive) · developer guide

Test what a job candidate can actually do before you interview them. You describe the role, add the skills it needs, and the AI drafts the questions. You check them, and every candidate sits an identical paper through a private link, with no account needed. The same tests work for your own team when you want to find out where the skills gaps are.

---

In one minute

1. LMS → Tests → New. Give the test a title and describe the role.
2. Add the skills, one row each: a difficulty, how many questions, and a format.
3. Fill all. Approved questions come from the bank first, and the AI writes the rest as drafts.
4. Read every draft, then Approve, Edit, Remove or Hide it.
5. Send. Enter the candidate's name, and their email if FreeITSM should send them the link.
6. Candidates shows their score, a score per skill and every answer they chose.

---

Describing the role

The first box is the most important one, because the AI reads it before it writes anything. "2nd line support" gets you generic questions. "A 2nd line analyst in a 40-person IT team supporting 900 users on Microsoft 365 and Intune, with some PowerShell and on-call one week in six" gets you questions about the job you are actually hiring for.

The test also has a time limit (the default comes from Settings; leave it empty for untimed) and an optional pass mark, which the result page uses to say Pass or Below the pass mark.

Skills, difficulty and format

Each skill is one row:

Skill	anything - Active Directory, Incident management, Networking, PowerShell
Difficulty	Beginner (the everyday task done the standard way), Intermediate (choosing between approaches, diagnosing a realistic fault), Advanced (trade-offs, unusual failures, design decisions)
Questions	1 to 20
Format	Multiple choice or Graded

The two formats

• Multiple choice - four answers, one right. Good for knowledge: "Which port does HTTPS use?"
• Graded - the scenario style of the old ITIL intermediate exams. Four answers that are all defensible but not equally good, worth 5, 3, 1 and 0 marks: the best action, a reasonable but weaker one, a poor one, and one that is wrong or harmful. Good for judgement: "47 users can't reach SharePoint and Microsoft shows no advisory. What do you do?"

Either way the candidate picks one answer, and they never see the marks. Every question counts the same towards the score: a graded answer worth 3 of 5 scores 60% for that question, just as a right multiple-choice answer scores 100%. That way a test with more graded questions isn't weighted towards them.

To use different marks, for example 4, 2, 1, 0, change them under LMS → Settings → Competency tests. New graded questions get the new marks, and questions already in the bank keep theirs (you can change any question's marks with Edit).

Filling a test

Fill all works through every skill that is short of questions:

1. It takes approved questions from the bank that match the skill, difficulty and format and aren't already on the test. Untick Use approved questions from the bank first if you want only new ones.
2. It asks the AI for whatever is still missing. The AI is told what is already in the bank on that skill, so it doesn't repeat it.

Each row shows how many it has, e.g. 3 of 5. Fill on a single row fills just that skill.

The AI is the one set up under LMS → Settings → LMS AI. Without it, Fill can only use the bank, and you can still write questions with Write or pick them with Bank.

Checking the questions

Everything the AI writes arrives as a draft. A test with a draft on it cannot be sent, so nothing reaches a candidate unchecked. On each question:

Button	What it does
Approve	it's right - it can be sent, and future tests can reuse it
Edit	fix the wording, the answers or the marks (changes it in the bank, for future tests)
Remove	takes it off this test only
Hide	retires it from the bank - it is never offered again (you can unhide it in the bank)

Approve all is there for when you have read them all. The AI is usually right and occasionally confidently wrong, so read them before you press it.

The question bank

LMS → Tests → Question bank holds every question ever written, whether by the AI or by you. You can:

• search by words in the question or its answers, the skill, or the role it was written for
• filter by skill, difficulty, format and status (draft, approved, hidden)
• tick several to approve or hide them together
• build a test by hand from the bank with Bank on the test

The bigger the checked bank, the cheaper and quicker each new test becomes.

Sending it to a candidate

Send asks for the candidate's name and, optionally, their email:

• With an email, and Email them the link ticked, FreeITSM sends it from your first mailbox that can send. The send is logged under the Training route.
• Either way, the link is shown once, with a Copy button. FreeITSM keeps only a scrambled fingerprint of it, so it can't show it again.
• Lost it? Press Link on the candidate's row to make a new one. The old one stops working. This only works until they press Start.

The link stays open for 7 days by default (Settings) until the candidate presses Start. Cancel withdraws it.

The candidate's paper is frozen at the moment you press Send: the questions, their answers in a shuffled order, and the marks. Editing or hiding a question afterwards never changes what they sit or how it's scored, and every candidate you send the test to gets an identical set of questions.

What the candidate sees

A page with the test's name, the number of questions, the time limit and four short rules, and a Start button. After Start:

• the questions, one after another, each with a single choice
• answers save as they go, so a dropped connection or a closed tab loses nothing
• a clock at the bottom, kept by the server. It keeps running if they close the page, and when it reaches zero, whatever they have answered is submitted.
• Submit when they're done (it warns about unanswered questions)

It works on a phone. Candidates never see the marks, the explanations or which answer is best, and by default they don't see their score either (Settings can change that).

Results

LMS → Tests → Candidates lists everyone you've sent a test to, with where they are: Not started, In progress, Finished, Link expired or Cancelled. Opening one shows:

• the overall score, and Pass / Below the pass mark if the test has a pass mark
• a score for each skill: where they're strong and where they'd need support
• when it was sent, started and finished, and the time taken
• every question with the answer they chose against the marks, and the explanation
• a box for your notes

Print gives a clean copy for the interview panel.

Privacy

Candidates are people outside your organisation, so their names, emails and scores are personal data:

• The whole feature sits behind its own permission, Competency tests, under Roles. It is not part of LMS manage, so somebody can run recruitment tests without managing courses, and the reverse.
• Finished, cancelled and expired candidates are deleted automatically 180 days later. Change this under LMS → Settings → Competency tests; 0 keeps them until you delete them.
• Delete on a candidate's row removes them and their result straight away.

Tell candidates how long you keep their results.

Settings

LMS → Settings → Competency tests:

Setting	Default
Links stay open for	7 days	until the candidate presses Start
Time limit for a new test	45 minutes	each test can change it; 0 = untimed
Marks for a graded question	5, 3, 1, 0	best first
Keep candidates' results for	180 days	0 = until you delete them
Show candidates their score	off	they never see answers or marks either way

Related pages

• LMS · AI authoring - the AI provider this uses
• Developer guide