An Adaptive Vigilance Task with Knowledge of Results

Wiener, Earl L. · 1974 · Human Factors The Journal of the Human Factors and Ergonomics Society

DOI: 10.1177/001872087401600401

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the feasibility of adaptive vigilance tasks, specifically examining how immediate knowledge of results (KR) influences performance when task difficulty is dynamically adjusted. Traditional vigilance research relies on fixed detection rates, which typically show a time-on-task decrement. Wiener proposes an alternative measure where the signal’s conspicuity (size) serves as an "adaptive variable," adjusting based on the subject’s recent performance to maintain a preset detection rate. The study aims to determine if this adaptive approach, combined with KR, can stabilize performance and provide a more sensitive measure of vigilance capability. The experiment involved four groups of undergraduate students performing a 48-minute, computer-controlled visual watchkeeping task. Subjects monitored a cathode ray tube for occasional signals consisting of two dots with slightly increased horizontal separation compared to non-signal stimuli. Two groups operated under fixed, non-adaptive conditions (one with KR, one without), while the other two groups operated under adaptive conditions (one with KR, one without). In the adaptive groups, the signal width was adjusted after every odd-numbered signal based on the subject’s detection rate over the previous eight signals. If the detection rate exceeded 75%, the signal became narrower (harder); if it fell below 75%, the signal widened (easier). The goal was to maintain a 75% detection rate. Results from the fixed groups confirmed traditional findings: the KR group maintained high detection rates throughout the vigil, whereas the NKR group exhibited a significant time-on-task decrement. The KR group also committed significantly fewer commissive errors (false alarms). In the adaptive groups, the variable signal width closely mirrored the traditional detection curves, validating the adaptive logic as a performance metric. The adaptive KR group successfully maintained a steady detection rate near the target, though slightly higher than the preset 75%, while the adaptive NKR group showed lower and less stable performance. Statistical analysis confirmed that KR significantly improved detection rates in both fixed and adaptive conditions. The study concludes that adaptive vigilance tasks are feasible and offer a valuable alternative to traditional fixed-parameter designs. By adjusting difficulty to maintain performance, researchers can measure the subtlety of signals an operator can reliably detect rather than just their detection rate. This approach has significant practical implications for training operators in fields requiring the detection of inconspicuous signals, such as industrial inspection, radar monitoring, or medical diagnostics. Adaptive techniques could allow for the training of novices to detect increasingly subtle flaws or the retraining of experienced monitors, providing a more precise assessment of skill and vigilance capacity.

Key finding

Immediate knowledge of results significantly improved detection rates and reduced false alarms, while adaptive signal adjustment successfully maintained steady detection performance over time.

Methodology

lab_experiment

Sample size: 64

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• signal detection theory