GESTA Gestalt Perception Test

A Rasch-homogenous test based on the hierarchic model perception to assess the construct field dependence.

Application
Assessment of the cognitive style of field (in)dependence (field articulation) through identification of a specific shape embedded in a pattern.

Theoretical background
GESTA is based on the concept of field dependence. This typology differentiates between field-dependent and field-independent people. Field-dependent people are influenced in their perception by surrounding stimuli, so that their perception is always influenced by the surrounding “field”. It is not just perception but the entire “mental apparatus”  (including cognitions and emotions) that is affected by this dependence on the “field”.  The concept of field dependence has developed in four phases.  The first tests were based on the theory of Witkin & Asch (1948). The first phase saw the introduction of the classic test for the assessment of field dependence, the Rod-and-Frame Test (RFT; Witkin, 1949). This was followed in the second stage of development by the test that is now most commonly used to measure field dependence: the Embedded Figures Test (EFT; Witkin et al, 1971).
The Gestalt Perception Test was drawn up on the basis of the hierarchical model. The aim is to measure the ability to deconstruct structures and reassemble them.  During development of the test, the explicit goal was to design what modern test theory would define as a unidimensional test that measures the construct of field dependence.

Administration
The task is to identify a specified shape (in the form of a house) within a pattern and to trace the outline of the house by marking the corners with the mouse.  The test consists of 30 items, all containing different patterns. The patterns are not just a jumble of lines; to varying degrees they have a “good” gestalt or shape. The shape that the respondent is looking for (the house) is always shown at the side, outside the pattern area of the item, and thus serves as a model.
For each item the respondent has 20 seconds to find the solution.  An item is classed as solved if the outline of the house is traced correctly (corners marked in the correct order) within the time limit.  A detailed instruction and practice phase precedes the test phase.

Test forms
There is one test form.

Scoring
The number of correctly solved items is calculated as the test score.

Reliability
Because of the validity of the Rasch model, the test can be assumed to have internal consistency. Cronbach’s Alpha varies in the different samples between r=0.89 and r=0.95. The split-half reliabilities are also very high, lying between r=0.83 and r=0.94.  

Validity
From the point of view of probabilistic test theory, reliability can be considered as given since all items measure the same ability dimension. The results on convergent validity are also especially important. The study by Hergovich and Kriechbaum (1996) shows that the correlation between the subtest “Analyze and Synthesize“ of the AID and GESTA is r=0.66. The correlation between EFT and GESTA is at r=0.51. The two tests EFT and GESTA are also found to correlate in similar ways with other variables that were investigated (extraversion, social desirability, intelligence).

Norms
Norms are available for a sample of N=443 individuals that is representative of the normal population in terms of age. Subsamples based on gender, age and education are also available. The data was gathered in two collection phases between 2002 and 2008 in the research laboratory of SCHUHFRIED GmbH, using a stratified quota sampling plan.