What Should You Do First? Cracking ASWB's Favorite Question Pattern

If you've taken a practice test, you already know the question type I mean. It gives you a client scenario, four answers that all sound reasonable, and asks something like "What should the social worker do first?" or "What is the most appropriate next step?"

These questions feel unfair because every answer choice describes something a competent social worker might actually do. The exam isn't testing whether you know good social work practice. It's testing whether you know the order of good social work practice. That's a learnable skill, not a guessing game.

Why these questions trip people up

Most people miss "first" questions for one of two reasons:

1. They pick the answer that solves the client's underlying problem, instead of the answer that's safe and appropriate right now.
2. They pick the answer that sounds the most clinically sophisticated, instead of the answer that's the most basic next move.

Both mistakes come from the same instinct: wanting to look competent by jumping to the deeper intervention. On the exam, that instinct gets you the wrong answer almost every time.

The priority hierarchy

When you see "first," "most immediate," "best initial response," or "next step," run the scenario through this order before you look at the answer choices:

1. Safety. Is there any risk of harm, to the client or to someone else, right now? If yes, that overrides everything else. Address safety before anything related to treatment planning, documentation, or long-term goals.
2. Assessment before intervention. If safety isn't in question, the next move is almost always to gather more information, not to act on information you don't have yet. You can't treat, refer, or intervene on a problem you haven't actually assessed.
3. The therapeutic relationship and informed consent. Before you dig into deep clinical work, the relationship itself needs tending: rapport, trust, clarifying confidentiality and its limits, making sure the client understands what's happening.
4. Direct intervention. Only after safety is addressed, the situation is assessed, and the relationship is established does the answer involving a specific technique, referral, or treatment plan become correct.
5. Documentation and follow-up. Usually last, even though it's professionally required. Documentation matters, but it's rarely the first thing you do in the moment with a client in front of you.

Memorize that order: safety, assessment, relationship, intervention, documentation. Most "first step" questions are really just asking you to locate where the scenario falls on that list.

A quick example of the pattern

Here's an illustration of how this works (not an actual exam item, just the pattern in action):

A client tells a social worker during an intake session that they've been having thoughts of ending their life but have no plan. What should the social worker do first?

A. Refer the client to a psychiatrist for medication evaluation B. Conduct a thorough suicide risk assessment C. Contact the client's emergency contact D. Develop a written safety plan with the client

All four are legitimate things that might happen eventually. But notice: C and A both jump ahead, assuming you already know the level of risk. D assumes you've already assessed and now you're intervening. B is the only answer that matches where this scenario actually sits: safety concern raised, but severity unknown. You assess before you act. The answer is B.

That's the whole trick. Find the step the scenario is actually sitting at, then pick the answer that matches that step, not the step after it.

How to eliminate answers that jump ahead

When you're stuck between two choices, ask these three questions:

• Does this answer assume information the social worker doesn't have yet? If a choice only makes sense after an assessment has already happened, it's probably wrong for a "first step" question.
• Does this answer skip the client's voice? Choices that unilaterally decide what's best for the client (without exploring, discussing, or assessing with them) are usually too far ahead, even if the underlying goal is reasonable.
• Does this answer solve the deeper issue, or does it respond to what's actually in front of you? The deeper issue might be real, but "first" questions are asking about the immediate next move, not the long-term plan.

A cheat sheet for common verbs in answer choices

If the answer choice starts with...	It's probably testing...	Usually correct when...
"Assess," "ask," "explore," "gather information about"	Assessment step	Safety isn't in question yet, and you don't have enough information
"Ensure safety," "ask directly about," "remove access to"	Safety step	There's any hint of risk to the client or others
"Explain confidentiality," "discuss the purpose of," "clarify the client's understanding of"	Relationship/consent step	Early in contact, before deeper disclosure or intervention
"Refer to," "develop a plan for," "implement," "teach"	Intervention step	Assessment and relationship-building have already happened in the scenario
"Document," "notify the supervisor," "report to"	Documentation/follow-up step	The scenario is past the point of direct client action

This isn't a perfect formula, and some questions will combine categories or test edge cases. But running the answer choices through this lens before you commit to an answer will catch most of the "too far ahead" mistakes.

Practice this deliberately, not passively

Don't just do practice questions and check if you got them right. For every "first step" or "next step" question you get wrong, go back and ask: which category did I think this was, and which category was it actually? Over time you'll notice patterns in your own thinking, usually a tendency to over-intervene or under-assess, and you can correct for that specific habit instead of generically "studying harder."

Takeaway

ASWB's "what should you do first" questions aren't trying to trick you with obscure content. They're testing whether you can locate a scenario on the sequence of safety, assessment, relationship, intervention, documentation, and pick the answer that matches that exact point, not the point after it. Learn that order cold, and these questions turn from the scariest part of the exam into some of the most predictable ones.