Cognitive strategies on solving spatial tasks: development of a dynamic spatial test in augmented reality

Within an interdisciplinary research project on cognitive strategies and trainingaspects of spatial ability, funded by the Austrian research fund FWF, we aredeveloping a new means of measuring spatial abilities in an ecologically valid way.Most of the conventional assessments measure spatial abilities in 2-dimensionalsettings; thus, they require 2D-3D-transformation processes that are often criticized.Developing a new spatial abilitiy measurement, we were using augmented realitytechnology and the software "Construct 3D", which allows for the projection ofvarious virtual objects into real space. Tasks can be viewed through electronicglasses (HMD's), and therefore be examined from different perspectives.Participants then actively have to construct solutions within virtual reality, usingspecial input devices.Empirical studies demonstrate, that performance on spatial tasks is often influencedby training and practice. Many of commonly found genderspecific differences inspatial abilities can - at least to some extent - be explained by differences inexperience with spatial tasks. Disregarding this fact can lead to an underestimation ofthe real potential of persons having less experience with spatial tasks (see Glueck,Kaufmann, Duenser & Steinbuegl, 2005). Thus, to gain more valid information on anindividual's spatial abilities, we are developing a dynamic test (containing a pretest,training phase and a post test), that measures the current performance level (pretest)as well as a person's potential for improvement after atraining phase.Spatial tasks used in the previous study are simple andcomplex figures consisting of single cubes. Thesefigures are presented within a 4x4x4 transparent grid(see figure 1). An item pool of 30 test items wasgenerated following construction rules based ontheoretical considerations. Those parameters aredimensionality, rotation axis and the position of therotation axis, complexity of the single objects, positionwithin the grid and number of rotations per item.