Table 1

Descriptive statistics for advice-taking by experimental condition

Low accountabilityHigh accountability
Type of Adviser: HumanAdvice Rationale AbsentMean = 5.56Mean = 5.64
St.D = 0.97St.D = 0.99
N = 33N = 36
Advice Rationale PresentMean = 6.77Mean = 7.86
St.D = 2.13St.D = 1.96
N = 31N = 29
Type of Adviser: AlgorithmAdvice Rationale AbsentMean = 5.67Mean = 6.13
St.D = 1.43St.D = 1.70
N = 33N = 30
Advice Rationale PresentMean = 7.52Mean = 7.76
St.D = 1.99St.D = 2.12
N = 27N = 29

Note(s): Every cell displays the mean, standard deviation and number of observations in the corresponding condition

Our dependent variable, advice-taking is measured on an 11-point Likert scale; higher values on the variable represent more advice-taking from the human/algorithmic adviser (depending on the type of adviser condition)

Type of adviser is manipulated between subjects on two levels: human/algorithm. In the human adviser condition, the advice comes from a human adviser. In the algorithmic adviser condition, the advice comes from an artificial intelligence tool

Advice rationale is manipulated between subjects on two levels: absent/present. In the advice rationale present (absent) condition, the advice from the human/algorithmic adviser [depending on the type of adviser condition] is (not) accompanied by an advice rationale

Accountability is manipulated between subjects on two levels: low/high. In the accountability high (low) condition, participants are informed that they are (not) being held accountable for their reported forecast

Source(s): Authors’ own work

or Create an Account

Close subscription notice
Close access options