NFC

For the Trustworthy Artificial Intelligence (TAI) class in my last spring term of the Computer Science (CS) masters we were assigned a group project. Related concepts are: trust, reliance, decisions, automation & bias.

I proposed an experiment (essentially: a quiz) where people have to answer a sequence of Multiple Choice Questions (MCQs). For each question the quiz offers a suggested answer with a supporting explanation. The suggestion may be correct or incorrect.

The experiment involves Cognitive Forcing Functions (CFFs). The control is no CFF: just the question, choices, and a suggestion. For half the questions: the person interacts with a CFF.

Anyway, DF from my group dug up this paper: [1] In the Introduction: TWMF introduction to Need for Cognition (NFC).

NFC was introduced in 1955 from the works of "AN EXPERIMENTAL INVESTIGATION OF NEED FOR COGNITION" [2].

References

[1]

Z. Buçinca, M. B. Malaya, and K. Z. Gajos, “To trust or to think: Cognitive forcing functions can reduce overreliance on AI in AI-assisted decision-making,” Proceedings of the ACM on human-computer interaction, vol. 5, no. CSCW1, pp. 1–21, 2021, doi: 10.1145/3449287.

[2]

A. R. Cohen, E. Stotland, and D. M. Wolfe, “An experimental investigation of need for cognition,” Journal of abnormal and social psychology, vol. 51, no. 2, pp. 291–294, 1955.

Acronyms

CS Computer Science 1

CFF Cognitive Forcing Function 1, 2, 3

MCQ Multiple Choice Question 1

NFC Need for Cognition 1, 2

TAI Trustworthy Artificial Intelligence 1