← Every exhibit
Why must algorithmic evaluation begin with task structure rather than system capabilities?
| Field | Question | Why it matters |
|---|---|---|
| Task character | Is the task perceived as objective, subjective, or mixed? | Perceived objectivity changes willingness to use algorithmic advice. |
| Advice source | Is the advice labelled as algorithmic, human, or unknown? | The source label can change adherence independently of the numerical advice. |
| Observed performance | Has the decision-maker seen the system make an error? | A visible mistake can change reliance, even when comparative performance remains stronger. |
| Comparison point | Is the advice compared with a person's estimate or with no alternative? | Appreciation can weaken when the user's own judgment is the direct competitor. |
| Expertise | Does the decision-maker have forecasting or domain expertise? | Expertise can change how strongly a person follows algorithmic advice. |
| Consequence | What happens if the advice is followed or ignored? | Reliance is a decision, not merely an attitude response. |
Swipe or scroll horizontally if the table is wider than your screen.
Reference & Evidence
Source: Table from this essay. Sources and interpretation are given in the article.
Related exhibits
-
The algorithm trust-by-task map
From the essay Algorithm trust changes with the task
-
Why must human-AI workflow design begin with discrete decisions rather than tools?
From the essay Human and machine should divide the work
-
Why must advertising measurement design begin with commercial decision thresholds?
From the essay More advertising observations do not guarantee better decisions