This experiment tests if additional information can suppress conclusions we may draw. We additionally test if non-monotonic operators such as normally have the same influence or not. In Summary we are interested if there are grades of nonmonotonicity (which can be supported).
Link 1: http://webexperiment.iig.uni-freiburg.de/human_deduction/
Link 2: http://webexperiment.iig.uni-freiburg.de/survey_cycles/
This experiment tests how we evaluate the consistency of 4 term series problems wrt. the way the information is sequentially presented (continuous, semi-continuous, and discontinuous). It tests a broad variety of problems.
Categorization: Four-term-series problems; Consistency; theory comparison
Link 1: http://webexperiment.iig.uni-freiburg.de/comparison2/
Link 2: http://webexperiment.iig.uni-freiburg.de/comparison2-en/
This experiment tests whether we prefer specific conclusions above others in causal reasoning. The results support the predictions of the preferred mental model theory.
Categorization: Three-term series problems; Reasoning with preferences; Causal Reasoning
This experiment is analogue to Experiment 2 but presented in German (30 students for we presented participants with verbal descriptions about cardinal directions of three stimuli. One third of the problems were determinate (i.e., allowing only for one arrangement), one third indeterminate (allowing for more than one possible arrangement), and one third with negated spatial information (implying as well an indeterminate arrangement). The results show an increase in reasoning difficulty (i.e., error rates from determinate over indeterminate to negated relations), but are slightly different to the first experiment in the accepted arrangements.
Classification: Arrangement of three objects; Verbal presentation (German); Large-scale relations
This experiment tests which conclusions are drawn wrt. the quantifiers Normally and Most. The findings indicate that the results can be explained by preference effects caused by a minimal mental model.
Categorization: Syllogistic Reasoning; Generalized Quantifiers; Most; Normally; Deduction Task with three sets.
Link 1: http://webexperiment.iig.uni-freiburg.de/ba_experiment/
Link 2: http://webexperiment.iig.uni-freiburg.de/eva2/
Logfiles Experiment 1:
Tasks Experiment 1:
Logfiles Experiment 2:
Tasks Experiment 2:
This experiment tests how we evaluate the consistency of three assertions (Ragni, Khemlani, & Johnson-Laird, 2013). Participants accept the assertion that elicits the preferred mental model and then the model that is not too distant from it (regarding the Levenshtein distance).
Classification: Syllogistic reasoning; Consistency reasoning; 3 sets
In this experiment (Ragni & Sonntag, 2012) participants have been presented with three quantified spatial premises. The premises have been connected by the connectors AND and XOR (exclusive or). Illusions (cp. Khemlani & Johnson-Laird, 2011) could be identified.
Classification: Syllogistic reasoning; spatial reasoning; spatial illusions
This experiment tests (in contrast to the deductive reasoning problems Experiment 3 in Ragni et al., 2007) what kind of assertions are accepted as consistent. The findings match the results from the previous experiment indicating that reasoning about consistency might be highly relevant for classical deductive tasks.
Classification: 5 terms; spatial reasoning; problems adapted from Ragni et al., 2007
Link 1: http://imodspace.iig.uni-freiburg.de/misc/spatial-syllog-consistency/
This experiment tests whether we can intuitively recognize if certain board combinations in Rush Hour are complicated or only look so. Participants had to calculate the number of moves (in their head).
Categorization: Planning problem; Cognitive Complexity; Decision problem
This experiment tests what kind of models participants construct for indeterminate descriptions of spatial syllogisms. The results indicate that the insertion strategies found for spatial reasoning are used for syllogistic reasoning as well.
Classification: Spatial reasoning; syllogistic reasoning; model generation task