MENU
C4I
Home
Center Overview
C4I
People
Objectives
Programs
Funding & Support
Industry Partners
Publications
Recordings
C4I
Events
News
Internet Conference
GMU C4I Center Seminar
Information order effects:
Examining the effect of Sequencing and Complexity
in a Long Series of Information
Dr. Ilean K. Keltz
Friday, April, 13, 2007 Science & Technology II Bldg, Room 320 2:00pm
ABSTRACT
This seminar reflects upon a study by Dr. Ilean Keltz. The study tested if the theory of anchoring & adjusting predicted the results for a long series of sequentially presented information when complexity, as defined by Hogarth & Einhorn (1992), and sequencing of directional information were manipulated. Hogarth & Einhorn's belief revision model predicted there would be a force towards primacy for a long evidence series, regardless of task variables because individuals anchor on early evidence and do not attend to recent information. This study tested two hypotheses. First, anchoring & adjusting does not always result in primacy when complexity and sequencing were manipulated. Second, that complexity and sequencing significantly affect belief revision through a mediator, mental effort.
Using Reserve Officer Training Corps Cadets, empirical evidence showed that the current belief revision model did not accurately predict a long series of sequentially presented information. Consistent with the first hypothesis, the interaction of the task variables order, complexity, and sequencing significantly impact belief revision and do not always result in a primacy effect for a long evidence series. In fact, when evidence was simple and grouped by direction, participants over adjusted to recent information resulting in a strong recency effect. When the evidence was grouped by direction and complex, participants did not over adjust to recent information resulting in a slight primacy effect. Additionally, when the information sequencing was mixed, there was a slight recency effect for simple evidence and no order effect for complex evidence. The mixed sequencing of evidence caused a reinterpretation of neutral information, which was designed not to favor a particular hypothesis.
The data however did not support the second hypothesis. Complexity and sequencing did not affect belief revision through the predicted mediator, mental effort, but the study recommends future research focus on measuring the mediator at the individual level, rather than globally as Hogarth & Einhorn's (1992) model assumes. This research found that anchoring & adjusting best describes the cognitive process of belief updating when evidence was sequenced by direction, just not necessarily in the way predicted by Hogarth & Einhorn. The study discussed engineering implications of designing systems to account for adjustment in a long evidence series. Additionally, the study recommended that future research focus more thoroughly on the task variables in a long evidence series and the effect that mixed sequencing had on the interpretation of neutral information.