Transfer of conditioned fear and avoidance: Concurrent measurement of arousal and operant responding

Rodríguez‐Valverde et al. (2021) taught 20 college students to form three 4-member equivalence classes. They used match-to-sample training until all subjects could pick the correct comparisons.

Next, one member of each class was paired with mild electric shock. The team then measured two things: how often the students pressed a key to avoid the shock, and how sweaty their skin got (a fear sign).

Avoidance spread perfectly. If A1 shocked you, you also dodged B1, C1, and D1 even though they were never paired with shock.

Skin conductance told a different story. Fear barely jumped to the other class members. The body 'knew' the difference even when the button press did not.

Vergason et al. (2020) also used an ABAB reversal, but in a zoo gift shop. Tokens changed staff greetings; here, shock changed key presses. Same design, very different worlds.

Wilder et al. (2025) asked whether medium-probability instructions work as well as high-p ones. Like the present study, they found one measure (cooperation) held up while another (speed) did not. Split outcomes seem common when we track more than one response.

Moxley (1989) cut agoraphobic avoidance with breathing drills. That clinical package beat relaxation and held for six years. The new lab work shows why: if you only kill the fear feeling, the avoidance may still ride the equivalence train. Treat both channels.

Your client who escapes all dogs after one bad bite may also avoid the park, the leash, and even the word 'puppy' once they form an equivalence class. Use that transfer: teach a safe dog, then probe the wider class for generalization. But do not trust heart-rate data alone; keep measuring the actual approach response. If the body still reacts, add in-vivo exposure or coping skills until the sweat drops too.