Carl Westin

76 Source bias effects Hum cfY Hum cfN Aut cfY Aut cfN 0 1 2 3 4 5 Number of participants (a) Advisories accepted ( N = 5) Hum cfY Hum cfN Aut cfY Aut cfN 100 80 60 40 20 0 Agreement rating (b) Agreement ( N = 4) Hum cfY Hum cfN Aut cfY Aut cfN 10 8 6 4 2 0 Response time (s) (c) Response time ( N = 4) Hum cfY Hum cfN Aut cfY Aut cfN 100 80 60 40 20 0 Difficulty rating (d) Difficulty ( N = 4) F IGURE 4-4: Barchart and scatterplots showing source and conformance results. 4-4-2 Questionnaire data Simulator questionnaire. Results from the simulator questionnaire showed that participants disagreed on two of the statements. Figure 4-5(a) shows responses to the statement: “I accepted resolution advisories even though I did not agree with them.” Two participants agreed with the statement, while three did not. This in- dicates that some participants accepted advisories even though they disagreed with them. Figure 4-5(b) shows responses to the statement: “I found the resolution advi- sory interrupting,” with two participants slightly disagreeing with the statement, and three slightly agreeing. Finally, responses to the statement in Figure 4-5(c) shows that all participants did not inspect the conflict before accepting an advisory. Source bias online questionnaire. Overall, SBQ questionnaire responses did not vary with advisory source. Rather, results indicate that participants had high trust in both advisory sources. Regardless of source condition, participants gen- erally agreed with the statements that the advisory system provided security, was dependable, reliable, had integrity, and could be trusted. Participants generally dis- agreed with the statements that the advisory system was harmful, deceptive, behaved