Job search
Thomas GIA Invitation: Prepare for Five Timed Tests and the Practice Instructions
Use current Thomas public instructions to distinguish response rules, practice examples and session scheduling without invented timings or score predictions.
A Thomas GIA invitation calls for preparation around five short timed tasks, each with its own instructions and practice. Start with Thomas’s public candidate guide and the linked example booklet, then reserve a quiet, uninterrupted session that fits your actual invitation. Learn the operation and the answer target for each task before adding time pressure.
Keep three things separate: reading the public examples beforehand, completing the practice presented inside your session, and taking the formal timed tasks. Familiarity with an example does not establish a score, an employer cutoff or permission to restart. Check unresolved access, timing or adjustment questions with the person who invited you before beginning.

Browse recent roles on LandOffer.ai while keeping this assessment’s invitation and preparation plan with the application it belongs to.
Confirm that this invitation is for GIA
Record the employer, role, assessment name, deadline and time zone from the genuine invitation. If it names several Thomas assessments, keep them separate. Thomas’s candidate guide describes Aptitude as timed; the same page’s Behaviour and Personality sections describe different, untimed activities. A statement about one assessment is not a rule for every task in the invitation.
GIA also appears under the Thomas Aptitude name. That naming connection does not make a third-party “intelligence test” practice page your official assessment route. Use the invitation to enter the employer’s process and the public materials for familiarization. Do not upload a private candidate link or identifier to a preparation service to ask whether its product matches.
Read the invitation’s instructions about completing the assessment at home or on the employer’s premises. Record the invitation’s support contact so a device, language or scheduling question can reach the person responsible before your reserved session. An invitation containing only a date leaves a practical question: does that date include a closing time and time zone? Resolve that before reserving the last available evening.
Recognize the five tasks without memorizing answers
Thomas’s Aptitude knowledge-base overview identifies Reasoning, Perceptual Speed, Number Speed & Accuracy, Word Meaning and Spatial Visualisation. These involve verbal information, matching, numerical relationships, vocabulary and mental manipulation of symbols. They are not a coding exercise or a review of your framework knowledge.
The 2025 candidate booklet shows the response mechanics. Reasoning uses a statement that disappears before the question; Perceptual Speed counts matching letter pairs; Number Speed & Accuracy compares distances from the remaining number. Word Meaning selects an odd word, while Spatial Visualisation distinguishes rotations from mirror images and counts matching pairs. Read the complete booklet rather than relying on isolated answer screenshots.
There is a material difference between two official descriptions. The knowledge-base summary describes Word Meaning as identifying two similar words, while the candidate booklet instructs you to select the odd word. Those actions are different. Do not guess which interface you will receive from a summary: follow your task’s actual demonstration and practice instructions, and clarify a discrepancy before starting its timed portion.
The preparation goal is to explain the requested operation and identify the response target. Use familiar examples to explain the rule in your own words, then check that your explanation still makes sense when the wording or symbols change. Avoid collecting supposed live questions or paying for claims of an employer’s exact answers.
Separate the task clocks from the session window
The current candidate guide gives a range of two to five minutes per sub-test and thirty to forty-five minutes overall, including instructions and practice. Thomas’s employer-facing GIA page summarizes the duration as forty-five minutes. Neither source gives you an employer-specific appointment or a custom accommodation timer.
Do not multiply five tasks by a guessed three-minute average and book only fifteen minutes. Nor should you add the full instruction-and-practice period a second time to the published total. Your calendar should allow the actual session plus a separate personal setup margin; label that margin as your planning choice.
The booklet describes eight practice examples before each task and recommends completing all five tasks consecutively in one sitting. Across five tasks that is forty practice examples, not forty minutes and not the number of scored questions. The booklet does not assign those forty examples a combined duration; the count cannot tell you how much calendar time their instructions and practice will take.
The instructions emphasize both speed and accuracy. This does not justify a guessed penalty formula, a universal guessing strategy or a claim that every candidate should answer a certain number of questions. Do not assign individual task timings from an unofficial simulator to the formal test. The timing displayed in your own session and any confirmed adjustment take precedence over a practice preset.
Worked case: Evan’s five-task instruction record
Evan is a fictional QA automation developer preparing for an invitation that explicitly names GIA. His example below is a completed paper analysis of public instructions, not a candidate account, practice attempt or assessment result. He writes a different response-control note for each task instead of copying its sample answers.
| Task | Misreading Evan identifies | Completed instruction-control decision |
|---|---|---|
| Reasoning | Looking back for the statement after moving on to the question | Before advancing, establish the relationship; then answer the question actually asked rather than a remembered adjective |
| Perceptual Speed | Selecting a letter because it looks familiar instead of reporting the match count | Read the matching convention and identify the count response; do not treat letter case alone as a mismatch |
| Number Speed & Accuracy | Selecting the largest number automatically | Identify the endpoints and compare each endpoint’s distance from the remaining value; select the endpoint the rule calls for |
| Word Meaning | Carrying the knowledge-base “two similar words” description into the booklet’s odd-word task | Mark the answer-action discrepancy and recheck the actual session demonstration before its timer starts |
| Spatial Visualisation | Assuming every rotated-looking pair is identical | Check whether rotation can align the symbols without reflection; determine what count the task requests |
His completed record separates three types of uncertainty. A misunderstood operation needs another reading of the official explanation. A response-target error means he knows the relationship but might click the wrong kind of answer. A platform question concerns the actual response control: whether the demonstration asks for a person, a number, an odd word or a count of matching pairs. More rapid clicking would not resolve any of those questions by itself.
Evan allocates twenty-five minutes beforehand: ten minutes to read the introduction and five-task descriptions, ten to write these five control notes, and five to list unresolved instructions. These are his invented preparation blocks, not Thomas’s test clocks. He has not completed the online practice or measured his speed.
He closes the paper review with this decision: “All five task names and response risks recorded; Word Meaning answer action must be confirmed in-session. Read the demonstration for every task, even one whose public example feels familiar. No readiness score assigned.” That is a specific preparation output without pretending to predict performance.
If Evan finds that a response demonstration is inaccessible or unclear, he contacts the inviter through the designated route. He does not start the timed task hoping he can return to the explanation later. Any approval, change or new instructions would be a separate record from his own preparation notes.

Worked case: Sofia chooses an uninterrupted window
Sofia’s independent fictional invitation names GIA only and gives a deadline of November 13, 2026, at 5:00 p.m. Eastern. She has a suitable laptop and a private room, but her calendar includes a short gap, a lunch window and a longer morning block. No real invitation, account or appointment was accessed for this example.
She compares the windows against the published overall range rather than adding five guessed task timers:
| Candidate window | Available time | Completed decision |
|---|---|---|
| November 11, 8:10–8:35 a.m. | 25 minutes | Reject: shorter than even the published thirty-minute overall lower estimate |
| November 11, noon–12:40 p.m. | 40 minutes | Reject for her plan: does not protect the forty-five-minute upper estimate, before any setup margin |
| November 12, 9:00–10:10 a.m. | 70 minutes | Reserve provisionally: ten minutes setup, forty-five for GIA including instructions/practice, fifteen uncommitted minutes afterward |
The chosen block finishes before her fictional deadline. The fifteen-minute margin is room in her own calendar, not extra permitted test time, a pause entitlement or a guarantee the platform will finish inside it. If the invitation later adds another assessment, she will schedule that component separately instead of assuming seventy minutes covers the bundle.
Sofia also completes a separate thirty-five-minute preparation plan for November 10: ten minutes reading the official introduction, fifteen reviewing all five task instructions, five checking her ordinary device setup, and five writing questions for the inviter. That is thirty-five minutes of preparation plus a seventy-minute calendar reservation: 105 minutes of planned availability, with only forty-five reserved for the estimated GIA session. It is not 105 minutes of testing.
Her device record says “laptop available; familiar external mouse available; power connection checked; private room reserved; invitation deadline recorded.” She leaves “employer-specific access instructions confirmed” open until she has read the actual invitation fully. Nothing in this fictional checklist establishes browser compatibility with an authenticated assessment.
Her final decision is: “Use the longer November 12 block, complete public-material review beforehand, and keep the lunch gap for ordinary work. Keep the twenty-five-minute gap for non-assessment preparation; reserve the formal session for the uninterrupted morning block.” She keeps the invitation contact available for unresolved issues without counting on a last-minute reply extending the deadline.
If Sofia needs an adjustment, her request must be discussed before relying on a changed arrangement. Her chosen room and mouse do not demonstrate that an accommodation has been approved. She would retain the response and revised instructions separately from the original invitation.
Check the device before beginning the assessment
Thomas’s candidate guide specifies a desktop or laptop, not a tablet or smartphone, and current Firefox, Edge, Chrome or Safari rather than Internet Explorer 11 or earlier. It also discusses background programs, pop-ups and JavaScript. Check the official instructions before starting; do not disable an employer’s managed security protections or change a device you do not control to force access.
The guide states that assessment material is loaded before the task begins. Treat that as the provider’s description, not a guarantee that every login, network interruption or final submission will work offline. We did not test an authenticated session. Keep a stable, authorized setup and resolve access problems before the timer where possible.
Use a mouse you are comfortable operating if that is your usual method. Check its movement and buttons in an ordinary application rather than opening a formal task as a hardware test. A quiet room also means arranging your own availability: close unrelated calls, tell household members when you need uninterrupted time, and avoid starting immediately before a commitment you cannot move.
Handle uncertainty without creating another attempt
If the instructions are unclear, return to the official explanation while still at the permitted instruction stage. Do not assume a browser back button, refresh or second login preserves progress. A public candidate guide is not proof of your employer’s reset policy or of an automatic extension after a technical problem.
If an error occurs, note the time, task stage and exact message without copying questions or answer choices. Tell the inviter whether the attempt appears active or completed and ask what to do next; do not create a second account or deliberately restart to test whether progress is retained. Describe whether the problem happened before launch, during practice or during a timed task; those are materially different states.
Keep completion evidence separate from interpretation of the result. An invitation marked complete does not establish a passing threshold, a hiring decision or the feedback your employer will release. Ask the hiring contact about the next stage instead of translating a practice percentage into a GIA percentile.
Your useful preparation record is the invitation, the five instruction-control notes, the uninterrupted session reservation and any confirmed arrangements. None substitutes for the actual demonstration when you begin. Browse recent roles on LandOffer.ai as you continue the rest of your job search.
LandOffer.ai publishes this guide. We reviewed public Thomas materials; no online practice, formal assessment, candidate account or scoring system was tested. Evan and Sofia are fictional planning cases, not candidate reports.


