Font Size:
Tracking the Wandering Mind. Exploring the possibilities of an eye tracking method for detecting periods of zoning out while reading.
Last modified: 2011-06-08
Abstract
While reading a text, a novel, an article, a scientific paper, one might have noticed that at certain moments one is actually not aware of what is written in the text, but rather concerned with different thoughts. Periods of thoughts which are independent from what one is doing at the moment are in general called mind wandering or ‘zoning out’ in the case of reading. There are several similar psychological concepts summarized by the term mind wandering, e.g. task-unrelated thoughts, task-unrelated images and thoughts, spontaneous conscious cognition or the probably most widely known phenomenon of ‘daydreaming’ [1]. Mind wandering is a common experience that makes up about 30% to 45% of our awake time [2]. Both as a venue of research into consciousness as well as due to its role in performance errors there is an increasing interest in methods of detecting mind wandering.
Currently, the method of choice for detecting periods of zoning out is experience sampling. Experience sampling is derived from so-called ‘descriptive experience sampling’, a method which is used to assess the contents of experience. The method consists of interrupting people in what they do and asking them about what was on their mind immediately before they were interrupted. In the case of studying mind wandering this means that the participants are instructed to perform a task, e.g. they are instructed to read a passage of text, and at random intervals they are asked whether in the moment before they were interrupted they were focusing on the text or on something unrelated to it. Due to it characteristics, this is a ‘probe-caught’ method. In contrast, using a so-called ‘self-caught’ method, the participants are instructed not only to perform a task, but also to simultaneously monitor what’s going on in their minds. If they become aware that they are actually not paying attention to the text anymore, they are supposed to press a button. Combining self-caught and probe-caught methods enables capturing the frequencies and the ending times of mind wandering episodes best.
Studies applying these methods revealed that the reading process can be divided in three basic phases:
* normal reading, i.e. reading with comprehension,
* zoning out with awareness, i.e. a periods of mind wandering that readers became aware of,
* zoning out without awareness, i.e. periods of mind wandering in which readers received a probe before they became aware of their mind wandering.
This demonstrates that the onset of a period of mind wandering usually occurs unnoticed [3].
Obviously, to study mind wandering with the help of experience sampling one has to interrupt and stop the process of mind wandering. A non-invasive method would therefore provide an advantage in studying fluctuations in attention and awareness. For this purpose experience sampling has been successfully integrated with different physiological measurements, like skin conductance, EEG, fMRI and several behavioral markers. Regarding reading, mind wandering has been investigated by combining experience sampling and eye tracking. Results showed that periods of zoning out and periods of normal reading differed in characteristics of gaze fixations, e.g. periods of probe-caught mindless reading (i.e. reading without the awareness of a lack of focus on the text) showed more off-text fixations than periods of normal or self-caught mindless reading [2].
Fixations, however, are difficult to track and analyze online and therefore only provide a weak method of assessing a reader’s state of attention during the reading process. The aim of the master thesis is to test the viability of alternative eye-tracking related parameters as markers of mind wandering. One specific hypothesis is that vergence increases during periods of mind wandering.
Possible conclusions will have to be drawn with care as eye tracking data is not merely influenced by the attentional state, but by various variables concerning text structure. Moreover, there may be differences in how readers monitor their degree of text comprehension and attention during reading.
So far eye-tracking data was collected from 6 subjects while reading long passages of text (sum of about 40 hours). Currently the prediction value and relationship of multiple eye-tracking parameters are being explored, with specific focus on pupil dilation, vergence and saccade speed parameters. These parameters are easy to track and process in real time. If they turn out to be useful, they will provide a good basis for investigating text comprehension, fluctuations of attention and consciousness.
[1] Smallwood, J. and Schooler, J.W. 2006. The Restless Mind. Psychological Bulletin. 132, 6, 946-958.
[2] Reichle, E.D., Reineberg, A.E., et al. 2010. Eye Movements During Mindless Reading. Psychological Science. 21, 9, 1300-1310.
[3] Schooler, J.W., Reichle, E.D., et al. 2004. Zoning Out while Reading: Evidence for Dissociations between Experience and Metaconsciousness. Thinking and seeing: visual metacognition in adults and children. D.T. Levin, ed. MIT Press. 203-226.
Currently, the method of choice for detecting periods of zoning out is experience sampling. Experience sampling is derived from so-called ‘descriptive experience sampling’, a method which is used to assess the contents of experience. The method consists of interrupting people in what they do and asking them about what was on their mind immediately before they were interrupted. In the case of studying mind wandering this means that the participants are instructed to perform a task, e.g. they are instructed to read a passage of text, and at random intervals they are asked whether in the moment before they were interrupted they were focusing on the text or on something unrelated to it. Due to it characteristics, this is a ‘probe-caught’ method. In contrast, using a so-called ‘self-caught’ method, the participants are instructed not only to perform a task, but also to simultaneously monitor what’s going on in their minds. If they become aware that they are actually not paying attention to the text anymore, they are supposed to press a button. Combining self-caught and probe-caught methods enables capturing the frequencies and the ending times of mind wandering episodes best.
Studies applying these methods revealed that the reading process can be divided in three basic phases:
* normal reading, i.e. reading with comprehension,
* zoning out with awareness, i.e. a periods of mind wandering that readers became aware of,
* zoning out without awareness, i.e. periods of mind wandering in which readers received a probe before they became aware of their mind wandering.
This demonstrates that the onset of a period of mind wandering usually occurs unnoticed [3].
Obviously, to study mind wandering with the help of experience sampling one has to interrupt and stop the process of mind wandering. A non-invasive method would therefore provide an advantage in studying fluctuations in attention and awareness. For this purpose experience sampling has been successfully integrated with different physiological measurements, like skin conductance, EEG, fMRI and several behavioral markers. Regarding reading, mind wandering has been investigated by combining experience sampling and eye tracking. Results showed that periods of zoning out and periods of normal reading differed in characteristics of gaze fixations, e.g. periods of probe-caught mindless reading (i.e. reading without the awareness of a lack of focus on the text) showed more off-text fixations than periods of normal or self-caught mindless reading [2].
Fixations, however, are difficult to track and analyze online and therefore only provide a weak method of assessing a reader’s state of attention during the reading process. The aim of the master thesis is to test the viability of alternative eye-tracking related parameters as markers of mind wandering. One specific hypothesis is that vergence increases during periods of mind wandering.
Possible conclusions will have to be drawn with care as eye tracking data is not merely influenced by the attentional state, but by various variables concerning text structure. Moreover, there may be differences in how readers monitor their degree of text comprehension and attention during reading.
So far eye-tracking data was collected from 6 subjects while reading long passages of text (sum of about 40 hours). Currently the prediction value and relationship of multiple eye-tracking parameters are being explored, with specific focus on pupil dilation, vergence and saccade speed parameters. These parameters are easy to track and process in real time. If they turn out to be useful, they will provide a good basis for investigating text comprehension, fluctuations of attention and consciousness.
[1] Smallwood, J. and Schooler, J.W. 2006. The Restless Mind. Psychological Bulletin. 132, 6, 946-958.
[2] Reichle, E.D., Reineberg, A.E., et al. 2010. Eye Movements During Mindless Reading. Psychological Science. 21, 9, 1300-1310.
[3] Schooler, J.W., Reichle, E.D., et al. 2004. Zoning Out while Reading: Evidence for Dissociations between Experience and Metaconsciousness. Thinking and seeing: visual metacognition in adults and children. D.T. Levin, ed. MIT Press. 203-226.