The effects of completing a task which requires the use of both automatic and controlled processes was investigated through a two-process experiment designed around a variation of the Stroop effect. Previous research found that, when performing certain tasks, response time is longer when an automatic process conflicts with a controlled process, in this instance reading interferes with naming the colour of ink a word is written in. These results reinforced a two-process theory of attention. In the current experiment, the nature of the words in which various colours of ink were printed was manipulated. The results further supported a two-process theory by showing that the nature of the words used did have a significant effect on response times.
If we consciously perceived the vast amount of stimuli available in daily life our senses would go into overload. Our brains have a restricted capacity and we only have the brainpower to attend to a limited amount of information. The conscious cognitive process of selective attention protects us from being overwhelmed by all the sensory signals impinging on our receptors by allocating processing resources where necessary. The conscious processing of information over which we can exert control (controlled processing) requires mental effort, drawing on the limited processing resources available to us in varying degrees depending on the task at hand, and can be easily interrupted. This was the stance assumed by Kahneman (as cited in Edgar, 2007) who put forward a ‘limited-capacity’ theory suggesting a general-purpose hypothetical mental structure with an upper limit in the amount of information it can deal with at any one moment. The responsibility of this processor is to analyse incoming stimuli and integrating it with information already within the memory, thus implying some information cannot be processed.
While this suggests the central processor divides its resource pool between competing ongoing tasks, there are instances where attention can be successfully divided under certain conditions. Through a succession of experiments Schneider and Shiffrin (as cited in Edgar, 2007) made a distinction between controlled and what they termed ‘automatic’ processes. Such processes require little or no mental resources and occur without conscious awareness, enabling our limited resources to be directed elsewhere allowing for some tasks to be done at the same time as others, thus preceding to the development of two-process theories. While automatic processing offers speed and economy of effort is has a notable disadvantage in that it can interfere with the conscious processing of information; a phenomenon that demonstrates said drawback is that termed the ‘Stroop effect’ (as cited in Edgar, 2007), which looks at what happens when we need to attend two conflicting signals.
In an experiment participants were required to identify the colours that were used to spell out the names of other colours, as swiftly as possible (the Stroop condition). In the alternate condition, in which participants tended to execute their response with greater ease, colour-neutral words were used (e.g. rat, grand, bolt, etc.). It would seem that the ability to read forms an automatic response, which during the Stroop experiments interfered with the controlled process of naming the colour of the ink. Through a variation of the Stroop effect, the present experiment investigated further the idea that automatic processing interferes with the information a person is consciously trying to attend. Rather than using colour words (e.g. red, purple etc.) the experimental condition employed colour-related words (e.g. blood, plum etc.) printed in a colour that was incongruent with the word; the control condition contained colour-neutral words (e.g. ledge, grade etc.).
The research hypothesis was that participants would take longer to complete the condition where the words were colour-related than the condition containing colour-neutral words. This is a one-tailed hypothesis. The null hypothesis was that there will be no difference in the times taken to complete the two conditions.
A within-participants design was employed. The independent variable (IV) was a list of coloured words which consisted of two conditions. Condition 1 (the experimental condition) was manipulated to compose of colour-related
words printed in a colour that was incongruent with the word. Condition 2 (the control condition) contained colour-neutral words. In both conditions the participant was required to say aloud the colour of the ink that each word was printed in. The dependant variable (DV) was the time taken to correctly identify the ink colour of each word within the colour-related word list. This was accurately measured to the nearest whole second by the researcher using a stopwatch. Controls were introduced to limit the effects of any possible confounding variables. To avoid possible practice effects of doing both conditions in the same order for each participant, the order in which the conditions were presented was counterbalanced.
Data on odd-numbered rows of the complete data set (1, 3, 5, etc.) are from participants who did condition 1 then condition 2, whereas data on even-numbered rows (2, 4, 6, etc.) are from participants who did condition 2 then condition 1 (A copy of the complete data set can be found in Appendix 1). Each colour-related word was used five times within condition 1; to avoid causing a practice effect the number of colour-neutral words used in condition 2 was the same. The same number of words was used in both conditions (thirty). The words used in the two conditions were matched for length. The different coloured inks were exactly the same hue for both conditions and presented in the same order. The instructions given to participants in both conditions were identical.
Of the twenty participants that took part in this experiment, 10 were male and 10 female, aged between 18 and 69. Sixteen of these participants were recruited from personnel at The Open University, or their family members and friends. The remaining four were selected from associates of the experimenter who had no knowledge of the hypothesis or the specific research sphere; all were informed that they would be taking part in a cognitive psychological experiment involving lists of coloured words, the resulting data would be used in a report to gain university course credits and that they could withdraw at any time. Each individual signed a consent form thereby obtaining informed consent.
The stimuli presented in each condition (a replica of which can be found in Appendix 2) consisted of a sheet of A4 paper on which 30 words, placed in two columns, were written in various colours of ink (red, blue, green, yellow, orange and purple). In the experimental condition (condition 1) colour-related words were printed in a coloured ink that was incongruent with the word (e.g. the word ‘grass’ printed in each of the above colours excluding its natural association, green). The words used were blood, sky, grass, lemon, carrot, and plum, each word appearing five times in each of its incongruent colours. The control condition (condition 2) contained colour-neutral words (blame, ledge, grade, career, plan and sty). Each word began with the same letter as its corresponding colour-related word (e.g. ‘Blood’ and ‘Blame’) as well as being identical in the number of characters; each of the words appeared in the list five times. The words were presented in no specific pattern, but the same in both conditions. Participants’ response time for each condition was measured with a stopwatch and rounded to the nearest whole second. Results were recorded on a pre-prepared response sheet. Verbatim instructions were issued to each participant.
Each participant was approached and asked if they would be willing to take part in an experiment that was investigating one aspect of cognitive psychology. They were informed of what participation would entail. If they agreed to take part, each participant signed a consent form (Appendix 3) and was reminded that they could withdraw at any time. The age, sex and order of conditions were then recorded on a pre-prepared response sheet before each participant was tested individually. The instructions for the experiment were read verbatim to each individual. They were told that they would be presented with two lists of words, one at a time, and that they should say out loud the colour of the ink that each word was written in as quickly as possible starting at the top left of the list working downwards and then proceed to the top of the right column.
To ensure they understood what was being asked of them they were shown an example sheet; once the participant confirmed that they understood what was required of them, either condition 1 or 2 was placed face down in front of them (the order of presentation alternating between participants to avoid a possible confounding variable). This information was also recorded on the response sheet. The paper was turned over and the stopwatch was started. When the task was complete, the time taken for each condition was recorded on the response sheet to the nearest second. The participant was then thanked, fully debriefed as to the aims of the experiment and given the opportunity to ask any questions they had.
The research hypothesis in this experiment was that participants will take longer to complete the condition where the words were colour-related than the condition containing colour-neutral words. The time in which it took for each participant to complete the required task was measured for each condition to the nearest second.
The results of the present experiment showed that it took longer to complete a task when it was required to attend to two conflicting signals at the same time indicating that automatic and controlled processes operate simultaneously. Response times in the condition where participants had to identify the colour of ink used for colour-related words were longer and therefore statistically significantly different from the condition where they had to identify colour-neutral word colours. This suggests that the process of reading interfered with participants’ ability to name the colour of ink each word was written in when the words were colour-related.
Stroop (as cited in Edgar, 2007) similarly demonstrated some of the costs associated with an interaction between automatic and controlled processes through the stroop effect experiment; namely that people tended to find it more difficult to respond with the colour of the ink a word was written in if the word itself described a colour, as opposed to a colour-neutral word. Stroop used his findings as evidence for a two-process theory of attention, indicating that automatic processing interfered with the information the participants were consciously trying to attend (controlled processing). If it were so that a general-purpose central processor divided its limited resource pool between competing ongoing tasks, as suggested by Kahneman (as cited in Edgar, 2007) it could be expected that in the present experiment the response times for the two conditions would be similar in value as the participants would find neither condition more difficult than the other.
However, this is not the case; on average participants did take longer to complete the experimental condition, therefore it can be assumed that they found it more difficult. This result implies that multiple pools of resources are present with regards to attention and that automatic and controlled processes operate concurrently. Numerous controls were put in place to ensure any possible confounding variables were at a minimum, however, there are factors that could contribute to these despite the practices put in place. Automatic processes can be influenced by individual strategies and so it may be that participants were able to exert extra control over their attention within the experiment. One method of doing this would be to focus their attention on the initial letter of each word, with the aim of ignoring the word itself as much as possible, allowing for a greater available resource pool which can then be applied to the identification of ink colours.
To combat this, the order in which participants completed each condition alternated, however this method has its limitations as it doesn’t guarantee the complete absence of a confounding variable. In conclusion, the results of the experiment reported here do support a two-process theory of attention. However, although the statistical results allowed for the rejection of the null hypothesis, it is possible that by performing the experiment on more than twenty participants, a greater difference in response times could be produced allowing for more solid support for the theory. Future studies conducted in this area should attempt to carry out experimental research on a larger sample of participants in order to strengthen the evidence and lessen the possibility of aforementioned confounding variables.
Edgar, G. (2007) Perception and attention. In D. Miell, A. Phoenix, & K. Thomas (Eds.), Mapping Psychology (2nd ed., pp.3-50). Milton Keynes: The Open University.