Appendix C. Exam day guide

Check the reference date. If a question specifies a reference date, evaluate it based on the statutes and public notices in effect on that date. If revisions were made after the verification date of this book, the exam is based on the provisions in effect.

Mark decisive facts first. In a scenario passage, mark ① what is being claimed (figures, rankings, efficacy, comparisons, prices), ② what the supporting data is and who produced it (is it an independent institution, what are the subjects and conditions), ③ who received what (manuscript fees, products, points, employment), ④ to whom, when, and through which medium it is sent (consent, transaction timing, nighttime, email), and ⑤ what data is entered where (purpose, consent, children, entrustment), and then read the options.

Remember scenario-reversal sets. Even if the same five options appear again in another question, The correct answer may be different. Do not choose an answer based on your memory of what you selected in an earlier question; judge again based on the factual circumstances of the current question.

Frequently used elimination rules.

1. Options that shift responsibility to AI services, agencies, or influencers are usually incorrect. Advertising responsibility lies with the business operator that ran the advertisement, and the duty of management and supervision remains even when entrusted.
2. Beware of options that cover up substantive issues by adding formalities, such as disclaimer phrases, 'AI-generated' labels, or citations of sources. Unsubstantiated claims, testimonial dialogue, and rights issues are not resolved by labels.
3. Look for options that import principles from a different situation. Examples include applying the nighttime exception for email to text messages, applying the transaction relationship exception to nighttime regulations, or applying the statistical special provision for pseudonymous information to personalized advertising.
4. Options that replace numbers with words ('Industry's highest level') or blur expressions ('Practically first') usually leave the same claim intact.
5. In measurement questions, options that compare time periods, terminate early based on interim results, compare metrics with different definitions, or switch solely based on statistical significance are common traps.

Tips by question format. For multiple-select questions, evaluate each option independently; for ordering questions, line them up by checking whether the output of the preceding step is the input to the next step; and for statement evaluation questions, verify the conclusion and the reasoning separately. In matching questions, there is one unused choice.

Time allocation. With 35 questions in 60 minutes, you have just over 1 minute and 40 seconds per question. Solve the questions you are certain about on your first pass, mark questions where time periods or numbers (6 months, 14 days, 15 days, 2 years, 9:00 PM to 8:00 AM, about 20 days) are confusing, and review them again in the final 15 minutes.