Date of Award

Summer 1997

Document Type


Degree Name

Doctor of Philosophy (PhD)



Committee Director

Mark Scerbo

Committee Member

Barry Clemson

Committee Member

Glynn Coates

Committee Member

Danielle McNamara


Two experiments examined whether people can develop a domain-general, nonlinear mental model when provided with an appropriate conceptual model. Both experiments presented causal loop models (i.e., variables connected in a circle by arrows representing causal relationships) as a conceptual model for positive feedback. Study 1 found that participants trained in causal loop modeling could accurately represent scenarios of low but not high complexity. Study 2 expanded on the design of Study 1 by varying the type of training and type of aid presented during testing. Participants received training with modeling instruction, training with cue-utilization instruction (i.e., participants were trained to represent scenarios as a list of bivariate relationships), general training, or no training. For a subset of testing scenarios, participants received either a modeling aid (i.e., a causal loop model) or a cueing aid (i.e., a list of bivariate relationships). Participants trained in modeling predicted the behavior of variables more accurately and quickly than those trained in cue utilization. In addition, participants provided with modeling aids during testing predicted system behavior more accurately and more quickly than those provided with cueing aids. In addition, spatial ability was found to correlate positively with accuracy scores. Overall, the results offer strong evidence that people can develop and successfully utilize a domain-general, nonlinear mental model.


In Copyright. URI: This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).