How to study for the EA exam

Use a small set of unfamiliar questions across the part’s published domains. Before checking answers, mark confidence and write the rule you used. A correct guess should enter the review list alongside wrong answers. Sort misses into missing knowledge, wrong rule selection, calculation, missed wording and time pressure. Those categories point to different remedies; repeating another random set does not necessarily address the cause.

Turn the result into a coverage map. Give each domain an initial status such as not studied, understood with notes, or explained without notes on fresh questions. These are editorial learning stages, not official score bands. The published domain counts help identify the breadth to cover, while the diagnostic identifies your personal gaps. A five-question result from one topic is too small to justify abandoning an entire domain.

Start with sessions you can actually protect. For example, four 60-minute weekday sessions and one two-hour weekend session provide six hours, not ten. Reserve part of that time for retrieval and review instead of allocating every minute to new content. If two hours are lost to work, move lower-priority new material and preserve a short review of previously learned rules. This is a planning method, not evidence that a particular number of hours guarantees a pass.

Use one part as the main focus while keeping a brief maintenance list for completed material where useful. Place difficult computation work in sessions with fewer interruptions. Batch short recall or error-log reviews into smaller gaps. At the end of each week, compare planned outputs with completed outputs: rules explained, fresh cases solved and recurring errors corrected. Time spent watching a lesson is useful context, but it is not itself evidence that the rule can be applied.

A useful error-log entry records the topic, the decision you made, the condition you missed and a retest task. “Review basis” is vague. “I used donor basis for a loss-valued gift without testing the sale against gift-date value” identifies a repairable mistake. Then solve a changed-price version without notes. If the same misconception survives, return to the rule explanation rather than memorize the previous answer.

Schedule later retrieval after a gap, using new numbers or a different story. Explain why a plausible alternative fails. Distinguish rules that need recall, such as a tested-year threshold, from procedures that need reconstruction, such as a basis rollforward. Keep the tested year visible on numerical notes. Mixing a current filing-season amount into materials for the exam’s specified tax year can turn a well-practised calculation into the wrong answer.

Move from untimed topic work to timed mixed sets once you can explain the method. Track accuracy, confidence, completion time and error type together. An improving score on repeated questions can reflect answer memory, so include unfamiliar material. Look for stability across domains and the ability to recover after a difficult question instead of relying on one unusually strong session.

The SEE passing standard is a scaled score, not a published raw percentage you can directly reproduce in a practice set. Do not label an arbitrary practice percentage “the IRS pass mark.” If a mock exposes several unresolved methods, use the domain map to target them and consider your scheduling options under the provider’s rules. The goal of the readiness review is an evidence-based study decision, not a guaranteed prediction.