[ CogSci Summaries home | UP | email ]

Trickett, S., Trafton, J., & Schunn, C. (2009). How do scientists respond to anomalies? different strategies used in basic and applied science. Topics in Cognitive Science, 1(4), 711--729.

	author =	(Trickett, S. and Trafton, J. and Schunn, C.),
	title =		(How Do Scientists Respond to Anomalies?DifferentStrategies Use in Basic and Applied Science),
	journal =	(Topics in Cognitive Science),
	year =		(2009),
	volume = 	(1),
	pages = 	(711--729)

Author of the summary: Author of Summary: Andrea Hills, 2011, andreagailhills@gmail.com

Cite this paper for:

1.0 Introduction

Anomaly- phenomenon that deviates from common form, inconsistency with what is expected (p 711).

Psychologists emphasize anomalies (p 711).

Kuhn: anomalies can lead to rethinking (p 711).

scientists who attend to anomalies are more likely to make progress (p12)

applied science: general science of phenomena to build predictive models of particular situations to solve practical problems and decision making (p 712).

basic science: develop, refine, and advance general theorectical scientific understanding (p 712).

anomalies are treated differently in basic and applied sciences (p 712).

in basic science: anomalies represent opportunity for discovery (p 712).

in applied science anomalies may be regarded as a nuisance (p 712).

basic scientists are more likely than not to attend to anomalies ( p712).

differences of strategies for handling anomalies depends more on the task being performed than the type of science (p 712).


1.1 Conceptual simulations

conceptual simulation: reasoning strategy, imagining a situation and mentally playing out the implications to see what happens, mentally "what if" reasoning (p 713).

three phases of conceptual simluation are initial representation, using mental operations to modify the representation to hypothetically alter the representation, and infering information from the mental result's validity (p 713).

conceptual simulations place heavy demand on working memory and require significant domain knowledge (p 714).

conceptual simulations are qualitative reasoning strategies that may be incomplete or imprecise but accomadate ambiguity (p 714).

conceptual simulations may be used as a method of evaluation that is faster and cheaper than other alternatives such as running experiments (p 714).

In basic science, inferences can be made about characteristics of conceptual simulations about underlyinh causes of theories (p 714).

constraints must be placed on the specifications of the simulations (p 714).

Applied scientists cannot use conceptual analysis for solving anomalies (p 714).

Scientists may mentally transform uncertainty that is mentally misrepresented (p 714).


1.2 Spatial transformations

Mental rotation is a form of spatial transformation to test spatial ability (p 714).

spatial transformation is a stratedgy used frequntly in complex visualizations (p 715).

Spatial transformation: spatial object is transformed from one mental state or location to another mental state or location (p 715).

Spatial transformations are analogous to transformations in physical space (p 715).

Spatial transformations include imagined objects, objects on-top-of another, modifying a mental imagines, mental rotation, animating a static image, transforming a 2 dimensional image into a 3 dimensional image, and making comparisons between different views (p 715).

Mentally transforming mental images can build a more complete picture of data for problem solving (p 715).

spatial transformations are discrete, individual manipulations (p 715).

conceptual simulation runs mental representations in a sequence that must have a starting representation to an inspectable ending representation (p. 716).

Pure spatial Transformation: mentally manipulating a mental representation so that it is a new representation. Compare: Comparison spatial transformation: comparing two images mentally (p. 716).

To understand anomalies, basic and applied scientists will use spatial transformations to compare different images and create new images to compare, sometimes by changing the display of images (p. 716).


2.0 Study 1

scientific thinking was studied by observing and recording the verbalization of scientists who performed their regular tasks (p. 716).

Using verbal protocols, authentic data of cognitive processes (Ericsson, K. A., & Simon, H. A. (1993). Protocol analysis: verbal reports as data (2nd ed.). Cambridge, MA: MIT Press.), the study identified anomalies and strategies for dealing with them (p. 716).

Two basic science and one applied science domains, involving expert (totals of ten years of experience) scientists with PhDs, trained to talk aloud using verbal protocols, were used for the study (p. 716).

To establish reliability, two separate people coded the verbal protocols (p. 717).

In basic science, verbal reference to whether something was anoamolous or expected, domain knowledge without explicit verbal reference, association with something already determined as anamolous, contrast to phenomenon to expected results, and scientific questioning of a feature were used to determine if something was analogous (p. 717).

Anomalies were identified by differing models or contrast between models and expectations (p. 718).

With good reliability, according to Trickett and Trafton (2007), anomalies across several utterances were encoded according to their reference to a new representation of a system or mechanism, transforming that representation spatially, in a hypothetical manner, and a result of the transformation (p. 718).

pure and comparison spatial transformation were categorized and coded when a participant mentally transformed one spatial object to another or location into another(p. 718).

In total, 17 anomalies were found in this expirement (p. 719).

This study controlled for base-line differences, timing before and after anomalies (p. 719).

More conceptual simulations occurred after the anomaly in general and with basic scientists (p.719).

the use of pure spatial transformations did not differ between the types of scientists, but were used after an anomaly more often (p. 720).

Data was examined for all sessions to examine the pattern (p. 720).

The results of the exaimined evidence suggests that conceptual simulation and pure spatial transformations were strategies used to respond to anomalies since they were used more after the anomaly was identified (p. 712).

comparison spatial transformations do not appear to be involved in responding to anomalies since its use was the same before as after the anomaly (p. 721).

the results of this study suggest that there are prodecural differences between basic and applied sciences responces to anomalies (p. 712).

Due to the low level of power, typical of this type of study, there was low statistical significance (p. 721).

Anomalies in basic sciences were identified when models and data did not match and an attempt to respond was made using concaptual simulation (p. 722).

In bsaic science, images were represented visually, manipulated, and analzed to deal with an anomaly, mentally playing out the data, to draw inferences about possible explanations (p. 723).

In applied sciences, discrapancies were attempted to be resolved using spatial transformations, mentally adding information to the model (p. 723).

Discrepancies between hypothesis and data are more likely to be resolved by conceptual simulation, whereas discrepancies between two sets of data are more likely to be resolved by spatial transformation (p. 723).

Particularily useful ways to solve tasks may perhaps require higher level skills an knowledge, gained by experts, in order to be implemented (p. 724).


3.0 Study 2

This study investigated the differences between experts and novices, getting them to complete the same task (p. 724).

Comparison spatial transformations require domain knowledge to know relevant comparison points (p. 724).

Each novice noticed at least one anomaly (p. 724).

Novices and experts did not differ in their use of either pure spacial transformations or comparison transformations before an anomaly (p. 725).

Timing was used as a within factor, and expertise was used as a between factor (p. 725).

No conceptual simulations were used by novices and very few were used by experts suggesting that they are not relevant to solving anomalies, althought novices may lack the skills to do it (p. 725).

experts used more pure spatial transformations than novices after anomalies (p. 725).

The interaction between timing and expertise was significant (p. 725).

Experts used many more times more pure spatial transformations after the anomaly, showing how it is an important expert strategy (p. 725).

marginally more comparison spatial transformations were used after the anomaly, while not much of an effect coming from expertise (p. 725).

Novices and experts used comparison spatial transformations after an anomaly (p. 725).

Experts used more pure spatial transformation but not conceptual simulations when they encountered an anomaly but novices used many more comparison transformations, suggesting novices merely compared anomalies than solved them, possibly because of a lack of knowledge (p. 725).



There has been three distinct problem solving strategies used in scientific reasoning: conceptual simulation, pure spatial transformation, and comparison spatial transformation (p. 727).

Expert basic scientists used conceptual simulations, expert applied scientists used pure spatial transformations, while novices lacked the skills to use pure spatial transformations (p. 727).

Novices focus more on data representations themselves, whereas experts move past the data (p. 727).

Scientists all mentally manipulated visual images (p. 727).

There may be different goals in science depending on the task (p. 727).

different domains may present different types of anomalies (p. 727).

The immediate task may determine the strategy used, while the domain and task may be interrelated (p. 727).

there is also a creative element to basic sciences that helps turn uncertainty into aproximation (p. 728).

Anomalies represent an important aspect of the scientific process (p. 728).

The distincation between basic and applied science has been validated by the findings of this study (p. 728).

Back to the Cognitive Science Summaries homepage
Cognitive Science Summaries Webmaster:
JimDavies (jim@jimdavies.org)