The review screen is your command center for the final-five routine. Sort by Unanswered first to clear all blanks, then work through Flagged items. For each flagged item, apply a lightning check: identify what the question asks, what must be true, and which choices violate that requirement. Keep guessing discipline consistent and reserve the last 45 to 60 seconds to ensure every item is answered.

Use back-solving on algebra by plugging middle values first to bracket the solution. For proportionality or linear models, verify with a unit check and a single test value. Try substitution with easy numbers for systems, or inspect slope and intercept without fully solving. In geometry, sketch minimally and test constraints. When time is nearly gone, estimate by comparing order of magnitude, sign, or endpoint behavior to eliminate three choices.

Think in filters rather than deep dives. For vocabulary-in-context, check the sentence before and after and pick the choice that maintains the author's tone and logical flow. For grammar, prefer the simpler, more direct option unless structure requires otherwise. When stuck between transitions like however versus therefore, interrogate the direction of contrast or cause. Fast elimination beats perfect certainty.

You budget for it by splitting each module into two passes. Pass 1 harvests easy points by answering everything you can in about 45 to 60 seconds while flagging longer items. Pass 2 works medium questions with structured methods. Track time every 5 to 7 questions and widen your skipping threshold if you fall behind. Practice with a visible countdown and build the habit of closing with five minutes left until it becomes reflexive.

Three effective drills include: Blank-Clear Sprints (2 minutes) where you open a mixed set with three scattered blanks and practice jumping and guessing with elimination; Near-Win Converter (3 minutes) where you finish two half-finished math and one Reading/Writing item using one fast method each; and Plausibility Gauntlet (2 to 3 minutes) where you scan five solved items and identify one reason each could be wrong such as scope word, sign, unit, or graph label errors.