Most interventions ask a client to do less of the thing that's causing trouble. Paradoxical interventions ask for more of it. The clinician tells the insomniac to stay awake, tells the procrastinator to schedule an hour of procrastinating, tells the couple locked in a nightly argument to have the argument on purpose at 8 p.m. It sounds like a trick, and on the licensing exam it's often written to sound like one. It isn't. It's a specific technique with a specific theory behind it, a specific home in the family therapy models, and a specific set of situations where using it would be a serious mistake.
Paradoxical work traces back to Milton Erickson and moves through the Mental Research Institute group in Palo Alto into strategic family therapy. On the exam, the names most likely to show up attached to it are Jay Haley and Cloé Madanes, whose work on the hidden function of symptoms and on family hierarchy sits at the center of the strategic model. The Milan group — Selvini Palazzoli and colleagues — built related techniques, including positive connotation and rituals prescribed to the whole family. Viktor Frankl's paradoxical intention, developed inside logotherapy, arrived along a separate path but rests on the same mechanism: a person who deliberately tries to produce the feared symptom stops fighting it, and the fight was doing most of the damage.
That lineage matters because a vignette can test placement rather than technique. A directive that seems designed to make the problem worse points toward the strategic model, and the question may be asking which approach is in use rather than what to do next. It's worth having the family therapy models and their signature interventions sorted before test day.
Haley's word for any between-session task is a directive, and most directives are straightforward: do this instead of that. Paradoxical directives are the subset the client may be expected to resist. They work through that resistance rather than around it, though they don't require it. A vignette may describe one without ever using the word.
Symptom prescription is the core form. The client is asked to produce, on schedule, the behavior or experience they came in wanting to get rid of. A man who worries he'll blush in meetings is asked to try to blush. A client whose worrying runs all day is asked to confine it to a scheduled half hour and to worry hard during it.
Restraining change is the mirror image. The clinician cautions the family to go slowly, wonders aloud whether they're ready for things to improve, or points out what the family would lose if the problem resolved. A family that's spent months pushing back against the clinician's optimism still has to push somewhere, and pushing against caution means arguing for change.
Ordeal therapy attaches an unwanted task to the symptom, so that having the symptom costs something. Pretending asks the client to fake the symptom rather than have it, which quietly replaces the involuntary version with a voluntary one and lets whatever the symptom was earning get earned some other way. Positive connotation and reframing aren't strictly paradoxical, but they travel with this family of techniques and get grouped with them on exams — recasting a mother's intrusiveness as devotion, or a teenager's defiance as an attempt to protect the parents from each other.
Two mechanisms are worth holding onto, because distractors are built out of confusing them.
The first is control. If a symptom is experienced as involuntary and the client produces it on command, it isn't involuntary anymore. Having it deliberately implies being able to not have it deliberately. The client's relationship to the symptom shifts from helplessness to authorship.
The second is the attempted solution. Strategic and MRI thinking holds that problems are maintained less by their original cause than by what everyone keeps doing to fix them. The partner who monitors, the parent who reminds, the client who fights the anxiety at 3 a.m. — the effort is the loop. Prescribing the symptom interrupts the effort. Sometimes the intervention isn't aimed at the symptom at all but at the solution attempt wrapped around it.
There's a third thing often described as the therapeutic double bind: the directive is structured so that either response moves something. Produce the symptom deliberately and the client has demonstrated a degree of control over what felt involuntary. Resist the directive and the client is resisting the symptom itself, which is the direction everyone wanted to go anyway. Neither outcome is a cure, and the technique isn't self-executing. But the bind is why it gets associated with resistance and with involuntary or reluctant clients — and why it gets confused with motivational interviewing's rolling with resistance, which is a stance, not a directive.
Underneath all of it is the strategic premise that the symptom is doing a job. It may be regulating distance between parents, holding a hierarchy in place, or securing attention the family has no other way to offer. That's also where first- and second-order change comes in. First-order change adjusts behavior inside the existing rules — the family does more of what it's already doing, harder. Second-order change alters the rules. Paradoxical work aims at the second, which is why it can look like it's ignoring the presenting problem while it works on the arrangement around it.
One contrast worth having ready: symptom prescription can look a lot like exposure, and the two get built into the same answer set. Exposure is a behavioral intervention aimed at new learning about feared cues and the responses attached to them, and it comes with a rationale the client agrees to in advance. Strategic symptom prescription aims to change the client's relationship to the symptom, to disrupt the attempted solution maintaining it, or both — and it doesn't depend on the client finding the logic persuasive. The instruction can read the same on the page. The model behind it is different, and that's usually what's being asked.
This is the part that turns a definition into a right answer. Paradoxical interventions require a symptom that's safe to produce. Prescribe nothing that can hurt someone.
That rules out suicidal behavior, self-harm, disordered eating, substance use, violence, and anything else where a worsening is dangerous rather than merely uncomfortable. It also rules out clinicians who haven't been trained in the technique. Guidelines for the safe use of paradoxical methods have been in the professional literature for decades, and the through-line is consistent: this approach calls for specific training and supervision, not improvisation.
The ethical questions are real too. A directive that lands partly because the client doesn't see the logic sits uneasily next to informed consent (NASW Code of Ethics 1.03) and next to the requirement to practice within one's demonstrated competence (1.04). It brushes against self-determination (1.02) as well — a directive is still the client's to accept or decline. None of that makes the technique unethical, and it doesn't actually depend on deception: a clinician can explain the rationale and give the directive anyway.
So when a vignette pairs a paradoxical option with a client at risk, a client in crisis, or a social worker with no training in the approach, the paradoxical option is a distractor no matter how well it fits the theory.
A social worker is seeing a couple. One partner describes spending an hour every night recounting and rechecking the household budget and can't stop. The other partner has begun sitting nearby, reminding them the numbers are fine and urging them to come to bed. The social worker, practicing from a strategic frame, wants to interrupt the pattern maintaining the behavior. Which intervention best fits that frame?
A. Ask the couple to keep a joint log of each checking episode and review it together in the next session
B. Direct the checking partner to set aside a fixed period each night for checking, and ask the other partner to stop offering reassurance
C. Explore each partner's family-of-origin history around money and financial security
A calls for monitoring — reasonable in a behavioral frame, but it adds to the surveillance already in the room rather than changing it. C is a legitimate piece of work in a psychodynamic or Bowenian approach and might matter later, but it doesn't touch the sequence. B does both things a strategic intervention is supposed to do: it makes the symptom deliberate and scheduled, and it removes the reassurance that's been feeding it. The reassurance is the attempted solution, and taking it out of circulation is at least half the intervention.
Best answer: B.
Notice what the question didn't include: any indication of risk. Add a line about escalating distress or a safety concern and B stops being the answer.
Paradoxical interventions reward studying that goes past the definition. Knowing the technique means prescribing the symptom handles a knowledge item. Knowing which model it belongs to, what it's actually aimed at, and when it's contraindicated is what handles a vignette.
Practice questions are the fastest way to find out which of those is in place. Get started with SWTP.
Right now, get SWTP's online practice exams at a reduced price. Just $39. Get additional savings when buying more than one exam at a time-- less than $30 per exam!
All tests reflect the new, August 2026 exam format.
Keep going till you pass. Extensions are free.
to receive our free study guide and get started!