“Watch out!”: Effects of instructed threat and avoidance on human free‐operant approach–avoidance behavior
A single verbal warning is enough to create steady human approach-avoidance loops, and raising the price of safety only delays the switch, not the beat.
01Research in Context
What this study did
Schlund et al. (2017) told adults to play a computer game. A red bar grew across the screen. If it hit the right edge, they lost money.
Players could press a key to move the bar left. Each press cost a few cents. The team raised the cost in some rounds. They tracked how often people let the threat build before they paid to reset it.
What they found
Just hearing "you may lose money" started stable see-saw behavior. People let the bar creep, then paid, then let it creep again.
When the key became more expensive, they waited longer before pressing. The back-and-forth rhythm stayed the same; only the switch point moved.
How this fits with other research
Barrett et al. (1987) showed that children who talk themselves through a task can slow their own button pressing. Schlund’s adults did the same trick after one brief warning, showing the skill lasts into adulthood.
Kendrick et al. (1981) found that telling people to "finish quickly" shortens their pauses on fixed-interval schedules. Schlund adds a threat twist: instructions still rule, even when money is on the line.
Rogers-Warren et al. (1976) used a visible clock to cut extra responses on avoidance schedules. The new study swaps the clock for a verbal rule and gets the same tidy pattern, proving either cue can trim wasteful work.
Why it matters
Your words alone can set off a client’s own approach-avoidance cycle. State the danger clearly, then let the client choose when to act. If the escape response is hard or costly, expect them to wait until the last safe moment. You do not need extra prompts; just adjust the cost or the rule and the cycle resets itself.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Before a risky task, give one clear rule like, "If you stay too long, you lose points." Then let the client choose when to exit; do not add extra reminders.
02At a glance
03Original abstract
Approach-avoidance paradigms create a competition between appetitive and aversive contingencies and are widely used in nonhuman research on anxiety. Here, we examined how instructions about threat and avoidance impact control by competing contingencies over human approach-avoidance behavior. Additionally, Experiment 1 examined the effects of threat magnitude (money loss amount) and avoidance cost (fixed ratio requirements), whereas Experiment 2 examined the effects of threat information (available, unavailable and inaccurate) on approach-avoidance. During the task, approach responding was modeled by reinforcing responding with money on a FR schedule. By performing an observing response, participants produced an escalating "threat meter". Instructions stated that the threat meter levels displayed the current probability of losing money, when in fact loss only occurred when the level reached the maximum. Instructions also stated pressing an avoidance button lowered the threat level. Overall, instructions produced cycles of approach and avoidance responding with transitions from approach to avoidance when threat was high and transitions back to approach after avoidance reduced threat. Experiment 1 revealed increasing avoidance cost, but not threat magnitude, shifted approach-avoidance transitions to higher threat levels and increased anxiety ratings, but did not influence the frequency of approach-avoidance cycles. Experiment 2 revealed when threat level information was available or absent earnings were high, but earnings decreased when inaccurate threat information was incompatible with contingencies. Our findings build on prior nonhuman and human approach-avoidance research by highlighting how instructed threat and avoidance can impact human AA behavior and self-reported anxiety.
Journal of the Experimental Analysis of Behavior, 2017 · doi:10.1002/jeab.238