Abstract
In this paper, we have studied the different types of information perception, such as color identification and reading speed. The research question is: “Is reading harder than identifying colors?” In order to investigate this problem, a quantitative study of the different levels of interference of reading or color identification in the Stroop test was conducted.
Research has demonstrated that when forced to perform an automated task, such as read the word ‘green’, it is more difficult when the word is printed in an alternative color, such as red, than when it is the same color as the word with which it is related, in this case, green. This has been used to establish evidence of the demands made upon the processing of information: Schneider and Schiffrin (1977) argue that there appears to be a limited capacity in how many tasks may be completed at the same time. This study utilizes a quantitative approach to gather research into whether the naming of colors is interfered with at a greater level by a reading task than the reading of colors is interfered with by non-congruent words. The results suggest that reading was more difficult than naming colors in the non-congruent tasks.
Introduction
Perception is the extent to which we can understand phenomena through our senses. However, we do not process everything that we see. Our brains need to be able to distinguish between different types of information in order to focus on different levels of importance. This process of sorting between what is perceived and the hierarchy in which we interpret this is attention (Edgar, 2007).
How attention functions has received interest from a number of researchers. Some, such as Kahneman (1973) argue that there are separate tasks that compete for a common pool of processing resources and so some things that we perceive are not processed. Other views suggest that we perceive a great deal, but simply do not process what is being perceived consciously (Scheider & Schiffrin, 1977). These are known as the early or late models of attentional processing (Edgar, 2007). Research examining the impact of these models of processing has illustrated that we more readily process some visual and aural phenomena than others (MacLeod, 1991; Edgar, 2007).
Research to examine the apparent hierarchy in which we process perception has demonstrated that there are systematic differences in our perceptual ability: some tasks can be completed effectively when they are completed together, whereas others seem to compete for the same processing power (Schneider & Shiffrin, 1977). This can be seen as offering support for the multiple-resource theory of attention. This can be also explained by the fact that some processes can be more successfully automated than others.
The classic manifestation of the Stroop test involves presenting participants with a list of words, some of which are the names describing colors (MacLeod, 1991). In the classic manifestation, two conditions are used: in the first condition, the words are all presented in the color to which the word refers, such as providing the word ‘green’ printed in green (Jensen & Rohwer, 1966). It is often observed that where the color of the word and the word are incongruous, participants take longer to read out the list of words (Edgar, 2007).
An interesting question that may be explored by this test is whether the recognition of colors or the recognition of words takes primacy in the processing hierarchy. It can be suggested that because the identification of colors is a behaviour that is learnt before the learning of reading (Duncan, 1980; Edgar, 2007). It may therefore be supposed that when placed in a position to choose between the two tasks, we are more able to identify a color than we are to read a word (Kahneman & Treisman, 1984).
This study will examine differences between three conditions: one where there is no difference between the word and the color it is printed in, and two incongruent conditions. In one incongruent condition the participants will be asked to specify the color and in the second condition they will be asked to read the word. It is anticipated that the test will be a one-tailed hypothesis because previous research indicates that the differences between the conditions should be reflected in a longer response time.
Before we create statistical hypotheses, we have to point out dependent and independent variables. An independent variable is a controlled variable. It can be actively changed by a researcher. A dependent variable is the variable (any mental phenomenon, characteristic), which changes are considered as a consequence of changes in the experimental exposure. In other words, this is a so called response variable, reflecting responses to experimental effects.
In our research work, the dependent variable is reaction time in milliseconds. The independent variable is a grouping variable – conditions (congruent/noncongruent word, congruent/noncongruent color).
The directional hypotheses for this study are as follows:
Reaction times will be fastest for congruent conditions
Reaction times will be moderate for non-congruent conditions when the participant needs to read the words.
Reaction times will be slower for non-congruent conditions when the participant needs to identify the color.
As a result of the research, the last two hypotheses were not supported. It is appeared that the reaction time is not moderate for non-congruent conditions when the participant needs to read the words. Also, reaction times were not slower for non-congruent conditions when the participant needs to identify the color.
Method
Design
This was a within-participants design. The independent variable was the extent to which the word is associated with a color. The dependent variable is the time that the participant took to react in each case. The order of each condition was alternated between participants to ensure that there was no systematic presentation of the conditions that would mean the participants would gain practice.
Participants
49 participants were recruited for this experiment. The participants were recruited as part of a convenience sample at a private university. 8 of the participants were male and 41 were female. The mean age of the participants was 25.92 years with a standard deviation of 8.379 years. The participants were naïve to the hypothesis.
Materials
The materials consisted of 4 words of colors and the test was conducted as part of a computer program. This helped maintain consistency between the participants. There was no time limit for the participants completing the test. Instructions were presented on screen, but a consent form was provided and signed by the participants.
Measures
The participants’ gender and age were recorded. The time taken to complete each condition was also recorded in seconds.
Procedure
The first step is cleaning and discarding data. Data cleansing is no rigging. When editing, encoding and copying the answers to questions researchers can detect extreme or unusual data. If a researcher tests a hypothesis that concerns numerical data, it is necessary to build a "stem and leaf" chart or a bar chart. This will clear the data and discard extreme values, which only distort the true picture. In addition, a preliminary analysis of the data should be accompanied by verification of assumptions about the distribution of the population. The process of data cleansing raises important ethical questions. Should we exclude certain data from the study? Of course, yes, if it turns out that the measurements were carried out correctly. Sometimes researchers have no choice - for example, a respondent may refuse further participation in the survey is not finished answering questions. The well thought-out study statistician should advance to formulate rules discarding data. In this case, we have detected all missing values and outliers and deleted those participants from the study (1 participant had missing data and one was an outlier).
The test was provided on a computer. Each test consisted of four words and was repeated four times, with a different test each time to ensure that the measures were reliable and less prone to chance. For the congruent condition, where the color of the word matched the word itself, the participants were required to read the words. For the second condition they were required to read the words when the color was incongruent with the words. For the third condition they were required to identify the color when it was incongruent with the words. In each case the time taken to read the colors was recorded.
The next step is a descriptive analysis. It consists of various measures of central tendency, frequency, impact of demographics on the test results, Pearson’s correlation coefficient and others. The following tables represent descriptive statistics for all quantitative variables, grouped by gender:
The correlation matrix was constructed in order to examine the impact of various factors on the test results:
It is appeared that there is a significant moderate positive relationship between CongruentRT and NoWord variables (r=0.56, p<0.001). Also, there is a significant weak positive relationship between NoWord and NoColor factors (r=0.424, p=0.003), a significant moderate positive relationship between NoColor and CongruentRT factors (r=0.549, p<0.001).
Now, we test the difference between the mean values of the three given conditions by gender, using T-test.
The t-test for independent samples indicated the following results:
There is no significant difference in the mean value of CongruentRT variable between males and females (t=-0.873, p=0.387)
There is no significant difference in the mean value of NoWord variable between males and females (t=-0.643, p=0.523)
There is no significant difference in the mean value of NoColor variable between males and females (t=1.212, p=0.232)
Finally, we used ANOVA in order to examine the reaction time in milliseconds depending on various conditions and gender. Two-way factorial ANOVA was performed in SPSS:
The ANOVA indicates a significant difference in reaction time between various conditions (F=4.218, p=0.017). However, gender and the intersection of both factors (gender and condition) do not significantly impact reaction time.
Results
The research hypothesis for this study is that it will take longer for participants to complete the test when the word and the color conditions are incongruent. It was further hypothesized that the time taken to identify colors in the non-congruent tests will take longer than the time to take the words in the non-congruent test.
Because this was a within-participants design, the results were tested using a chi-square test (X2(49)=15.11; p=0.001). This suggested that the differences between the means were significant. This confirms the hypotheses.
The t-test for independent samples indicated no significant difference in the mean value of CongruentRT, NoWord, NoColor variables between males and females.
The ANOVA indicates a significant difference in reaction time between various conditions. However, gender and the intersection of both factors (gender and condition) do not significantly impact reaction time.
Discussion
The results suggest that when presented with congruent words and colors, response time is the fastest. The results also indicate that the identification of the color is faster than the processing of a word. This suggests that when reading, the color of the word interferes more with the participants’ reading speed than the reading of the word interferes with the identification of the color. The results show that automatic efforts can intrude on thinking even where the participant is making a consistent effort to think about another task. The results lend some support that when reading, the meaning of the word is processed immediately rather than being consequential.
The implications of the test may be that this effect indicates the differences between the points at which the performance of these different tasks became automatic varied between the identification of color and the reading task (Gopher, Weil & Siegel, 1989). This may be expected, given that color recognition occurs at an earlier stage than reading. However, this presents an interesting implication that rather than a task being rendered automatically, it competes with other tasks according to a hierarchy of use and some tasks are perhaps given precedence (Edgar, 2007). Although the process of either reading the word or identifying the color is automatic, they cannot be entirely ignored and the fact that these processes are automatic interferes with the test.
Although this study holds some validity, the sample size of 49 is relatively small and there is a significant skew towards female participants, with a relatively limited age range. This limits the extent to which the results may be deemed valid because it is unclear if the observed phenomena are limited to this population sample. However, the implications of the results may be that there is a hierarchy of automation in our processing capacity, which supports research conducted by Das (1973). It may be an interesting avenue of further research to establish how far these results compare with other tasks, such as where aural processing and visual processing are combined, or where numbers, colors and words are combined. Although it may be hypothesized that the order in which the processes are automated will result in the hierarchy of results, this may not be the case when the processing of different faculties is compared, such as comparing aural and visual processing (Edgar, 2007). It should also be noted that the interference of one process on another does not necessarily indicate the primacy of one task over another; the capacity of the color of a word to interfere with the reading of the word may simply indicate that reading the word takes greater effort than color recognition (Meier & Kane, 2013).
However, this may be challenged by the fact that in color recognition, the participant has to recall the color name, whereas in the reading task all the information is presented and thus the reading task may be completed without knowledge of the words involved. This may suggest simply that the reading task requires more attention even when automatic and thus the potential of the colors to interfere is higher. Further research may explore these implications.
References
Duncan, J. (1980). The demonstration of capacity limitation. Cognitive Psychology, 12(1), 75-96.
Das, J. P. (1973). Structure of cognitive abilities: Evidence for simultaneous and successive processing. Journal of Educational Psychology, 65(1), 103-108.
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.
Gopher, D., Weil, M., & Siegel, D. (1989). Practice under changing priorities: An approach to the training of complex skills. Acta Psychologica, 71(1), 147-177.
Jensen, A. R., & Rohwer, W. D. (1966). The Stroop color-word test: A review.Acta Psychologica, 25, 36-93.
Kahneman, D. (1973). Attention and Effort. Englewood Cliffs, NJ: Prentice-Hall.
Kahneman, D., & Treisman, A. (1984). Changing views of attention and automaticity. Varieties of Attention, 1, 29-61.
MacLeod, C. M. (1991). Half a century of research on the Stroop effect: an integrative review. Psychological Bulletin, 109(2), 163-203.
Meier, M. E., & Kane, M. J. (2013). Working memory capacity and Stroop interference: Global versus local indices of executive control. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(3), 748-759.
Schneider, W., & Shiffrin, R. M. (1977). Controlled and automatic human Psychological Review, 84(1), 1-66
Wickens, C. D., & Carswell, C. M. (1995). The proximity compatibility principle: its psychological foundation and relevance to display design. Human Factors: The Journal of the Human Factors and Ergonomics Society, 37(3), 473-494.