Multi Location Training: How to Run Assessments Across Venues and Cities

Multi location training is a course delivered at more than one venue, city or country. Running its assessments well means a learner in one room sits the same blueprint, rules, timing and scoring as a learner anywhere else, and you can show that the results are comparable.

This guide is for training providers who sell one course and test learners at several venues, not for a retail or restaurant chain rolling out onboarding. As of October 2026, searches for “multi location training” mostly return guides for franchise, hospitality and manufacturing operators, and they rarely cover how the test itself is run. This post adds a standardization framework, a venue readiness gate, a day-of-exam checklist and a simple score comparison between venues.

Key takeaways

• Standardize the blueprint, item bank version, rules, timing and pass mark everywhere, and let venues vary only in room, local staff and local start time.
• Qualify each venue before exam day by testing bandwidth, devices and the exact exam, because the ITC/ATP guidelines call for testing the delivery system before operational use.
• Define the exam window in one reference time and decide in writing what happens when a connection fails, before anyone sits the test.
• Compare venues after every session using mean, spread, pass rate and item difficulty, and treat a gap as a reason to investigate, not proof that one venue is at fault.

What does standardizing assessments mean in multi location training?

Standardizing across venues means every learner meets the same test content, instructions, time limit, conditions and scoring rules, so a score reflects what the learner knows and not where they sat. Differences unrelated to the skill become noise in the results.

The ITC/ATP Guidelines for Technology-Based Assessment, published in 2022 by the International Test Commission and the Association of Test Publishers, name this problem. They describe construct-irrelevant variance as occurring “when item or test scores reflect factors the test was not intended to measure.” A weak Wi-Fi corner can add exactly that kind of variance. Guideline 3.14 adds that “Testing should be conducted under appropriate environmental conditions, and outcomes should be interpreted in light of those conditions.” So control the conditions you can, and record the ones you cannot.

Which parts of the assessment must be identical at every venue?

The blueprint, item bank version, rules, time limit, pass mark, scoring and accommodations policy must be identical everywhere. Room layout, local staff and local clock time may differ. Decide this split once, in writing, and log any other difference as a deviation.

The table shows each layer of a standard, what stays fixed, what can vary, and the evidence to keep for each venue.

Two rows deserve extra care. A single form keeps scores directly comparable, and the item bank versioning rules keep every venue on the same item versions. A single pass mark means one documented decision, as in the worked panel example in how to set cut scores for professional training courses, not a local pass mark per venue.

How do you check that a venue is ready before exam day?

Test each venue one to two weeks ahead, using its own network, devices and room with the real exam, and record a pass or fail. The ITC/ATP guidelines (3.7) say the delivery system should be “evaluated and confirmed to be accurate before operational testing.”

Use this readiness table as a gate: a venue hosts no live session until every row passes.

The guidelines (comments on 3.1) suggest identifying “dead zones and other areas of lower bandwidth capacity that may not be suitable for testing.” They also say (3.2) that requirements such as “bandwidth required per individual test taker, hardware, and software needs” should be provided in advance, so send your spec to every venue in writing.

Pro tip: Rehearse At Full Load. Run the connection test at the same hour as the real exam, with the same number of devices joining at the same moment, because a lone laptop on a quiet network proves nothing about a full room at peak time.

How should invigilators be briefed so every venue runs the same rules?

Give every invigilator the same written briefing script, read aloud at each session, plus a one-page rules sheet and an incident log. Same rules, same words and same records matter more than using the same person.

The script should cover identity checks, what is allowed on the desk, the exam length, what to do if a screen freezes, how questions are handled, what counts as an incident and how to close the session. Guideline 3.12 adds that “Contact staff should be trained to answer routine questions and escalate those requiring more technical assistance.”

Supervision level should match the stakes. Most in-person providers sit at the room-invigilation level described in proctoring for training providers, so the aim here is identical supervision at each venue, not more of it.

Observed practical exams need more. A 2022 study by Homer in Advances in Health Sciences Education modeled 442 administrations of an 18-station clinical exam and found examiners fell into more stringent “hawk” and more lenient “dove” groups, and that adjusting for stringency changed modeled pass rates. The setting is medical, but the lesson carries to any rated skills test: if different people mark at different venues, calibrate them on the same sample performances before the session.

How do you handle time zones, scheduling windows and daylight saving?

Define the exam window once in a single reference time such as UTC, show each venue its local equivalent, and confirm both before sending invitations. Typing local times by hand into each venue’s schedule invites missed or early starts when daylight saving changes.

Time zone rules do change. The IANA Time Zone Database is updated as governments alter offsets and daylight saving rules, and its page for release 2026e, dated September 29, 2026, notes that Manitoba moves to permanent UTC-5 on October 31, 2026. Check your scheduling system against a current release before sessions that fall near a change.

Choose how venues relate in time and state it in the learner notice:

• Same moment: every venue starts together, which limits sharing between venues but may mean an awkward local hour.
• Same local time: every venue starts at, say, 10:00 local, which suits learners, but early venues can pass information to later ones.
• Staggered by load: starts are spread to protect bandwidth, an option the ITC/ATP guidelines list when capacity is short.

What is your fallback when the connection or equipment fails?

Decide the fallback before exam day: who can pause a session, how a learner resumes, when a retake is granted and what gets recorded. The ITC/ATP guidelines (3.9) say the platform should let test takers “resume the test (where they stopped or as close to it as possible) after a service disruption or a planned break.”

The table lists common failures, the first response and the rule to settle in advance.

Log whatever happens. Guideline 3.8 says “System performance should be monitored throughout testing,” and your incident log is the human version of that. Flag attempts that used a fallback so they can be checked when you compare venues.

What is the day-of-exam checklist for each venue?

Give each venue lead the same four-stage checklist: set up 90 minutes before, admit and brief learners, supervise the session, then close and report within 24 hours. Used identically in every city, it leaves a comparable record whoever runs the session.

Step 1: Set up (90 minutes before). Check seats against registrations, power and spare devices. Run the connection test and a sample exam on every device. Confirm the invigilator and support contact are present, then post the rules sheet and set out accommodations.

Step 2: Admit and brief (30 minutes before). Check ID against the register and collect phones and notes as your rules require. Read the briefing script word for word, including the failure procedure, and start only at the agreed reference time.

Step 3: Supervise (during the exam). Log every question, interruption and device issue with the clock time. Apply the pause and resume rules exactly as written, with no local exceptions, and keep the support contact informed of delays.

Step 4: Close and report (within 24 hours). Confirm every attempt submitted and note any that did not. Send the incident log, headcount and deviations to the central owner, with the venue and room recorded against each attempt.

How do you consolidate results from every venue in one place?

Give every attempt the same fields: learner ID, venue, room, invigilator, form, session ID, start and end times, mode (screen or paper) and any deviation code. One shared record lets you compare venues and apply one pass mark without chasing spreadsheets.

SimpliAssess’s features page lists Item Banking, Test Management, Examinee Management, Schedule Management and Analytics & Reporting, and its training companies page says assessments can expand to “multiple platforms, locations and users.” Its published pages do not describe venue-level roles, time zone handling or offline delivery, so ask any platform you consider, including SimpliAssess, how venue, room and invigilator are captured per attempt and whether they export.

How do you check that scores are comparable across multi location training venues?

Compare each venue with the others using average score, spread, pass rate and an approximate interval around the gap. A gap larger than chance is a prompt to investigate causes such as equipment, instructions, supervision or learner mix, not proof that the venue is at fault.

The table shows an illustrative check. The numbers are invented for teaching and come from no real provider. Each venue’s mean is compared with the other two venues combined, and the interval is a rough normal approximation (gap plus or minus 1.96 times its standard error).

Venue A looks higher, but its interval includes zero, so the data cannot separate it from chance. Venue C sits about 7.7 points below the others and its interval stays below zero, which warrants a look. With only 12 learners the estimate is wide, so one session is a signal to check, not a verdict.

Then go to item level. Using the item difficulty method in item analysis for trainers, compare each item’s percent correct by venue. If one image-based item is answered correctly by 83% at Venues A and B but 33% at Venue C, suspect a rendering fault or a different instruction before suspecting the learners. The ITC/ATP guidelines (comment on 1.12) say “If multiple devices or platforms are used, studies should ensure the TEIs operate similarly,” where TEIs are technology-enhanced items.

When venues have large samples, a formal differential item functioning analysis can test whether an item behaves differently for two groups. Holland and Thayer (ETS, 1986) describe the Mantel-Haenszel procedure as a method that may be used to study differential item functioning in two groups of examinees. Small venues cannot support it, so rely on the simple checks and pool sessions over time.

As starting triggers, our own editorial rules and not a published standard, investigate when the interval excludes zero, or when pass rates differ by 15 points or more with at least 10 learners per venue. Then check likely causes in order: learner mix (compare pre-course scores), equipment and rendering, instructions and timing, supervision, and marking.

Pro tip: Tag The Room And The Person. Record the room and the invigilator against each attempt, not only the city, because a city average can hide a single room or a single marker whose conditions differ from the rest.

What should you do next?

Start with one high-stakes exam and two venues. Write a one-page standard from the layer table, run the readiness gate, use the same briefing script and checklist at both, and compare the first session before adding venues.

If past results carry a venue field, run the comparison on your last cohort first. It takes about an hour and shows whether you already have a venue problem.

If you also run the sessions across sites, SimpliTrain, a separate training management product from the same company as SimpliAssess, has a page on its multi-location training management system. It describes centers with their own admin teams and reporting by center and in total, and lists assessment scores per location as a metric, but it does not describe how exams are delivered across venues, so ask about that directly.

Frequently asked questions

What is multi location training?

Multi location training is training delivered at more than one site, such as venues in several cities or countries. For a training provider it means the same course and, ideally, the same assessments run in each place. The challenge is keeping content, rules, timing and scoring identical so results from different venues can be compared.

How do you keep exams consistent across multiple venues?

Use one blueprint, one item bank version and one pass mark, then standardize the rules, timing and briefing script. Qualify each venue before exam day, use the same day-of checklist everywhere, and tag each attempt with its venue and room. After each session, compare venue results to spot differences worth investigating.

Do all venues need to sit the same exam form?

Not always. A single form is simplest and keeps scores directly comparable. Parallel forms built to the same blueprint can reduce item exposure, but they need checking, and strict score interchangeability may call for psychometric help. Whichever you choose, record the form on every attempt and compare results by form as well as by venue.

How many learners per venue do you need to compare results?

There is no universal minimum. The precision of a venue average improves with the square root of the number of learners. In the example above, a 12-learner venue with a standard deviation near 11 has a standard error of about 3 points on its average, so only large gaps stand out. Pool sessions over time for small venues.

How do you handle time zones for an online exam window?

Set the window in one reference time such as UTC and display each venue’s local time from that single source. Decide whether venues start at the same moment, the same local time or staggered times, and tell learners which applies. Check your scheduling tool against a current time zone database before sessions near a daylight saving change.

What should happen if one venue’s internet fails during an exam?

Follow a rule written in advance: pause the session, contact the support person, and resume the learners where they stopped once the connection returns. The ITC/ATP guidelines say platforms should let test takers resume after a service disruption. Log the delay and flag affected attempts so they can be reviewed when you compare venues.

Should every venue use the same invigilator?

No. Venues need the same rules, not the same person. Give each invigilator the same briefing script, rules sheet and incident log, and a trained contact for technical questions. For observed practical exams, calibrate markers on the same sample performances, because different markers can be more or less strict.