Attention and Implicit Memory

Research Areas: 

AIM is a subproject of MEHME and investigates priming as a part of the human long-term memory in conjunction with attention. As stimuli we use visual real-world scenes. The aim of the project is, to found out influences of priming to attention, and vice versa. Further a question is, how far the effect of implicit memory processes to attention can be distinguished from explicit memory processes to attention?


Methods and Research Questions: 

For priming is a very brief presentation of visual informations sufficient, because it works without our awareness or attention. But there is evidence, that priming on his part has influence on attention and our eye movements. This leads to the question of which kind the influence is and weather it could be used as an alternative measurement for priming?

Priming is part of the human long-term memory. It facilitates the processing of a given visual information based on a prior encounter with that same or a related visual information. Usually for priming is a very brief presentation of a stimulus sufficient. However in previous studies with a longer exposure time of a stimulus is a repetition effect observable. In repeated scenes subjects make fewer but longer fixations given a fixed viewing time. So they sample fewer regions of a previously viewed photographed scene than for a novel one. This effect may be used as an indirect, implicit measurement for memory.
Further an effect occurs, if subjects are presented scenes and later the same scene are repeated with a subtle alteration (e.g. an object was removed or added). This results in an increase of the fixations of the altered region.
Both effects suggest that eye movement and attention is influenced by implicit processes. Resulting from this, the research question is which memory processes facilitate these effects? Is it only depending on priming or is it interconnected with semantic memory?

In the AIM project we work with photographs of real-world scenes. Currently the stimuli set consists of 40 photographs each in an original and a manipulated version. In the latter version an object or person is added or removed. The photographs are presented in three blocks. For this purpose they are divided in three groups: novel, repeated and manipulated. During the experiment novel scenes are seen once, repeated scenes are seen once in each block (overall three times) and manipulated scenes are seen once in an original version in the first two blocks and then in a manipulated version in the third block. Across participants additional each scene is randomized, so that each scene is viewed equally often as novel, repeated and manipulated. Eye movements are recorded during all three blocks.
This setting allows us to evaluate the visual stimuli in conjunction with the eye tracking data to investigate the differences in the eye movements between the different types of scenes and to make predictions about the impact of priming on attention. Supplementary a second condition with a explicit task (to activate explicit memory processes) make a comparison between these two processes possible.



The expected outcome of the project is, that priming not only works in the minute of a very brief presented stimuli. Rather it is a system working most of the time without our awareness and influences our perception. An additional outcome of the project is a corpus of eye movement data. It contains gaze data of first encounter with real-world scenes as well as data of repeated viewing.
Currently we are conducting a study with the described stimuli set. First results show that the effect in manipulated scenes is stronger if the alteration destroys the semantically consistence of the scene. That suggests that semantic knowledge is involved in the process.