Table 3. Results of confirmatory factor analysis.

Latent variable	Observed variable	ƛ	α	CR	AVE
Trust (TRU)			0.70	0.71	0.45
	TRU1: I trust my partly automated car to maintain speed and distance to the car ahead.	0.66
	TRU2: I trust my partly automated car to keep the car centered in the lane.	0.64
	TRU3: I can trust my partly automated car.	0.72
	TRU4: I feel hesitant about activating the partly automated car mode from time to time (reverse-coded).	Omitted due to factor loading < 0.60
	TRU5: I am unwilling to hand over control to my partly automated car from time to time (reverse-coded).	Omitted due to factor loading < 0.60
	TRU6: I always know when my car is in partly automated driving mode.	Omitted due to factor loading < 0.60
	TRU7: I engage in other activities while driving my partly automated car.	Omitted due to factor loading < 0.60 and merged with question of construct ‘NDRTE’ (non-driving related task engagement)’
	TRU8: I monitor the performance of my partly automated car most of the time.	Omitted due to factor loading < 0.60 and merged with questions of construct ‘driver engagement’
Perceived safety (PS)			0.81	0.82	0.60
	PS1: I feel safe most of the time.	0.78
	PS2: I feel relaxed most of the time.	0.88
	PS3: I feel anxious most of the time (reverse-coded).	0.67
	PS4: I feel bored most of the time.	Omitted due to factor loading < 0.60
	PS5: I am concerned about my general safety most of the time (reverse-coded).	Omitted due to factor loading < 0.60
	PS6: I entrust the safety of a close relative to my partly automated car.	Omitted due to factor loading < 0.60
Driver engagement (DE)			0.70	0.70	0.44
	DE1: My partly automated car always reminds me to take back full control.	0.64
	DE2: My partly automated car always reminds me to keep my hands on the steering wheel.	0.68
	DE3: My partly automated car helps me to keep using it in the manner as advised by the manual.	0.66
	DE4: The surrounding elements (other road users, road borders, marks) detected by my partly automated car are always clear to me.	Omitted due to factor loading < 0.60
	DE5: I monitor the performance of my partly automated car most of the time.	Omitted due to factor loading < 0.60
Non-driving related task engagement (NDRTE)			0.71	0.71	0.55
	NDRTE1: I engage in other activities while driving my partly automated car.	0.75
	NDRTE2: I use my partly automated car because it helps me to use my time for other activities unrelated to driving.	0.74
Performance expectancy (PE)			0.72	0.73	0.56
	PE1: I use my partly automated car because it helps me to reach my destination more comfortably.	0.74
	PE2: I use my partly automated car because it makes driving more pleasurable.	0.77
	PE3: I use my partly automated car because it helps me to use my time for other activities unrelated to driving.	Omitted due to factor loading < 0.60 and merged with question of construct ‘NDRTE (non-driving related task engagement)’
	PE4: I use my partly automated car because it helps me to reach my destination more safely.	Omitted due to factor loading < 0.60
Automation use (AU)			0.68	0.71	0.52
	AU1: Please indicate how often you activate ACC in your car.	0.80
	AU2: Please indicate how often you use your partly automated car with speed and steering support.	0.68
	AU3: Please indicate how often you activate LKA in your car.	Omitted due to factor loading < 0.60

Note that ƛ are the factor loadings, which are interpreted as correlation coefficients for the relationship between the questionnaire items and their underlying constructs.

α is the Cronbach’s alpha reliability coefficient, which is a measure for the internal consistency of a latent construct assuming that the correlations between the questionnaire items underlying a latent construct are equal.

Composite reliability is also a measure for the internal consistency of a latent construct, using the varying factor loadings of the questionnaire items on their underlying constructs and their error variance as input for the calculation.

AVE is the average variance extracted that is accounted in the latent construct among the questionnaire items underlying a latent construct.