Skip to main content
Integrity6 min read

Review flagged sessions

Work the three integrity lists on the Needs a look card: misconduct records, incomplete attempts and interviews without a camera. A review queue, not a verdict.

Before you start

  • Admin access to your college's Expert Hire console, signed in with your institutional Google or Microsoft account.
  • A cohort that has attempted interviews. The lists are empty until they have.
  • The date range for the term you are reviewing.
Steps

7 steps, in order.

1

Set the date range, then read the Needs a look card

The Dashboard opens when you sign in. Set the date range first, because everything on the page follows it and the default range can make an active cohort look idle. The card is headed Needs a look and carries four rows: Suspected misconduct, counted against the interviews in range, Incomplete attempts, meaning interviews completed in under 10 minutes, Interviews without camera, and Low scoring candidates. The first three are the integrity rows and are what this guide covers. Low scoring candidates is a scoring signal and belongs with the readiness work instead.

Suspected Misconduct, Interviews Without Camera and Incomplete Attempts, next to the cohort averages.

2

Open Misconduct Records and read the table for what it is

Suspected misconduct opens Misconduct Records, described on the page as students flagged for suspicious activity during interviews. The table carries name, registration number, branch, company, role, score, date and the events recorded on that session. Treat it as a review queue, not a verdict. Events have innocent explanations: a notification arriving, a dropped connection, a second monitor.

  • Nothing in this list rejects a student, fails an attempt or changes a score. It exists so a person can look.
3

Filter by event type and look for repetition

Above the table sit a name search, an All Event Types dropdown and Reset filters. Narrow to one kind of event and read down the list. A single incident on one session tells you very little. The same event type appearing across several of one student's sessions is the pattern worth a conversation. When nothing matches, the page says so plainly with No misconduct records found.

  • Filter to one event type at a time. Mixing event types in one read is how a device fault gets mistaken for a habit.
4

Check Incomplete Attempts, which is a duration list

Incomplete Attempts lists interviews completed in under 10 minutes. The table holds name, registration number, company, role, round, score, duration and date, with a name search and a registration number search above it. Most short attempts are technical: a microphone that never came up, a connection that dropped. Check for a pattern by student before you assume disengagement.

5

Open the camera-off list from the interview table

Interviews without camera opens Mock Interviews with the camera-off filter already applied. The table is the full cohort interview list, with registration number, name, company, role, round type, score, date and status, filterable on every column. Sort it by branch before you conclude anything: a count that sits almost entirely inside one branch is a lab or a briefing to fix, not a set of students to talk to. A student who sits a round without a camera also loses the behavioural analysis on their own report.

Every attempt across the cohort, filterable on any column, with the camera-off filter one of them.

  • Run the camera-off filter branch by branch. It is the fastest way to catch a lab full of machines with no working webcam.
6

Open the row before you conclude anything

Any row opens the full session and the events recorded on it, and on the interview side that is the same report the student sees: score, strengths, focus areas, the action plan and the question by question summary. Read the report next to the flag. A short attempt with a real score behind it is a different story from a short attempt with nothing in it.

7

Turn the filtered list into a conversation

The interview table has Send Interview Notification for emailing the students you have just filtered down to, and Download CSV for taking the view into a meeting. Both are how a flagged list stops being a number on a dashboard. Name the students you are actually going to speak to before you leave the page, because a filtered view is not saved and rebuilding it a week later takes longer than reading it did.

What you get

At the end of this guide.

Three review queues reached from one dashboard card: misconduct, incomplete attempts and camera-off interviews.

A misconduct table with the event type on every row, filterable to one kind of event.

A list of every interview that ended in under 10 minutes, with the duration on the row.

The cohort interview table with the camera-off filter applied, so you can see whether it clusters by branch.

The full session and the student's own report behind every flagged row.

A notification action and a CSV export for turning a filtered list into a conversation.

Common mistakes

What goes wrong, and the fix.

Reading a misconduct row as proof of cheating.

It is a review queue, not a verdict. A notification, a dropped connection or a second monitor all produce events.

Acting on a single incident.

Look for repetition across sessions. One event on one session is noise.

Assuming a short attempt means the student gave up.

Most attempts under 10 minutes are technical. Check for a pattern by student before you draw a conclusion.

Reading the counts without setting the date range.

Everything on the dashboard follows the range, and the default can make an active cohort look idle.

FAQ

Questions people ask on this one.

Still stuck? Talk to us and we will walk through your setup.

Read next

Related guides.

Placement operations

Track how ready your batch is

Read the admin dashboard the way it is built to be read: averages first, then the risk cards, then the outcome bands, then the individual students behind them.

6 min read
Campus

Onboard a student batch

Send the admin list, get your placement team signed in, invite the batch, then set the companies and roles your students practise against.

6 min read
Integrity

Turn on proctoring

Web proctoring for low-stakes rounds, the desktop app for strict ones, no extra credits either way, and every flag reviewed by a person.

6 min read
All guides

Try this in your own workspace.

The 30-day trial includes 75 Hire Credits, which covers a resume batch and a first round of interviews. Nothing charged for 30 days.

30-day free trial with 75 Hire Credits. Card required, nothing charged for 30 days.