Proceed with caution: using verbal protocol analysis to measure situation awareness.

Verbal protocol analysis (VPA) is often used to elicit information about the cognitive processes of operators as it provides rich data and can be used in naturalistic settings. Recently VPA has been used to investigate the acquisition and maintenance of situation awareness (SA), and to make comparisons between groups despite a lack of research regarding the efficacy of using VPA for this purpose. This train simulator experiment investigated whether VPA can effectively measure SA. Novice and expert participants were recorded on an audio device while talking aloud throughout the trials and their verbalisations were transcribed verbatim. A coding scheme developed from the transcripts was used to code the verbalisations. Results did not support the use of VPA as a measure of SA but did show that VPA detected differences in SA errors. Potential reasons for the conflicting findings between this experiment and those conducted by other researchers are discussed. Practitioner summary: This paper examined the validity of verbal protocol analysis (VPA) as a situation awareness (SA) measure. A repeated measures experiment was conducted using a train simulator. Normal VPA did not detect changes in SA but a measure of errors did. Caution should be used when using VPA to measure SA. Abbreviations: LETSSA: low-event task subjective situation awareness technique; SA: situation awareness; SAGAT: situation awareness global assessment technique; TPD: train performance display; VPA: verbal protocol analysis.