Free GSOA practice test: 30 questions with answers

Key judgements with explicit confidence and evidence let decision-makers judge the recommendation. Raw screenshots, cover colours, and author biographies do not convey confidence.

A clear bottom line and stated assumptions help readers act. Raw results without synthesis, unsourced claims, and decorative charts reduce usefulness and credibility.

Applying classification and minimising content protects sensitive information. Full working files, public publication, and unencrypted channels expose it.

A forecast with a time horizon and probability language is the appropriate form for future-oriented judgements. Certainty, current facts, and unsupported opinion do not fit.

Scenario analysis explores alternative plausible futures and their drivers. Single articles, employee counts, and ignoring uncertainty do not explore alternatives.

Predefined indicators with thresholds provide early warning. Waiting for obviousness, headline tracking, and random unrelated data do not give early signals.

Ranking by relevance, reliability, and time makes the investigation efficient. Alphabetical order, excitement, and unstructured pursuit waste effort.

Correcting the record and re-evaluating related conclusions keeps the analysis sound. Keeping flawed conclusions, hiding errors, and deleting notes compound the problem.

Documenting searches, sources, dates, and reasoning enables reproduction and review. Memory-only processes, conclusions alone, and context-free screenshots prevent reproduction.

Separating research from personal accounts reduces exposure and attribution risk. Personal accounts, public passwords, and unauthorised use of others' accounts are unsafe or unlawful.

Minimising and handling personal data according to privacy rules and purpose is required. Publishing, selling, and assuming public means no protection are unacceptable.

Deception to access restricted areas can breach terms or law, so the request should be escalated. Creating or using identities without authority is not appropriate.

Transparent methods, sources, and dates let the analyst judge reliability. Colours, advertisements, and domain length are weak indicators of reliability.

Reliability is topic-specific, so claims outside expertise need corroboration. Full acceptance, complete rejection, and ignoring the topic are not proportionate.

Defining the intelligence requirement and the decision it supports focuses collection and analysis. Account lists, extensions, and slide formats come after the requirement is clear.

The intelligence cycle structures intelligence work from requirements through dissemination. The other cycles describe different domains.

Open-source intelligence is publicly available information obtained lawfully and ethically. Classified signals, stolen credentials, and private records are not open sources.

Web archives store historical captures of pages, which reveal earlier content. Current code and recent snippets do not show the past, and asking the owner may not be possible or appropriate.

Metadata and reverse image search can reveal the original and context, which tests the claim. Appearance, pixel counts, and self-confirmation do not verify the image.

Several independent clues that agree give stronger geolocation confidence. A single clue, colour balance, or posting time alone is weak evidence.

Consistent resolution, angle, and timing separate real change from artefacts of collection. Mixed seasons, colour-only comparisons, and captions do not establish change reliably.

Additional dates and seasonal consideration distinguish construction from vegetation change. Immediate conclusions, ignoring, and unverified publication are unsound.

An assessment with intent, capability, confidence, and indicators is actionable. Raw posts, malware copies, and context-free maps do not support decisions.

Attribution requires evidence, alternatives, and confidence to be useful and honest. Definite answers without evidence, name-based guesses, and refusal are not appropriate.

Independent corroboration is the strongest approach. Profile photos, follower counts, and name matches alone are weak and can be faked or coincidental.

Timing and content similarity relative to normal behaviour indicate coordination. Colours, emojis, and backgrounds are weak or irrelevant signals.

Clarifying the decision ensures effort is useful and relevant. Collecting everything, refusing without dialogue, and generic reports do not serve the requirement.

Tailoring length, format, and messages to the audience and decision makes intelligence usable. Raw dumps, personal templates, and unprioritised findings reduce usability.

Filings list officers, directors, and major shareholders, which reveal control. Slogans, personal blogs, and plant lists do not reveal control.

Shared addresses and directors are a lead for investigation, not proof of wrongdoing. Declaring fraud, dismissing the link, and assuming good management are unsupported.