Introduction

The roots of Stroop's (1935) research are evident 50 years earlier in James McKeen Cattell's (1886) doctoral project. Cattell reported that objects and colors took longer to name aloud than the corresponding words took to read aloud. That is, saying "squirrel" to a picture of a squirrel or saying "red" to a patch of color was slower than saying "squirrel" to the word squirrel or saying "red" to the word red. Surprisingly, no one thought to combine colors (objects) and words until Stroop (1935). It is also interesting to note that Stroop developed his color word-task less from an interest in stimulus-naming time than from an interest in interference between conflicting processes. He was concerned with how best to explain interference. Since 1935, there have been hundreds of experiments conducted on the Stroop effect. C. M. MacLeod (1991) provides an overview of the findings, proposed theories and some well-grounded empirical facts (some of which are listed at the end of this discussion).

There are a number of theories regarding the Stroop effect; however, the discussion here will focus on attention as it applies to two types of cognitive processes: automatic versus controlled processes. Although the definition of attention is surprisingly elusive, most scientists agree about several key points. Attention is thought to be selective--focused on one object or train of thought at a time (James, 1890); it is thought to be a limited resource, such that there is a limited amount of attention available for cognitive processes (Posner, 1975); and attention appears to be intimately tied to learning (Shiffrin & Schneider, 1977). Theories of attention focus on two types of cognitive processes: automatic and controlled. Traditionally, it has been assumed that AUTOMATIC processes are involuntary, do not require attentional resources, and are fast ; whereas, CONTROLLED processes are voluntary, require attention for their execution, and are relatively slow (Cohen, Dunbar & McClelland, 1990). This presumed dichotomy between the use of attentional resources and process type is illustrated in Figure 1. From this traditional view point, the Stroop effect is a relatively clear example of competition between an automatic (color word reading) and a controlled (ink color naming) processes. Thus, naming the ink color draws more heavily on attentional resources than does reading the irrelevant word, and, consequently, it takes longer to name the ink color than it does to read the color word.

The problem with applying the traditional all-or-none automaticity assumption to the Stroop effect is that it implies that reading is automatic. Reading, like most tasks that eventually become "automatic," develops gradually with practice, and it is clear that even automatic tasks like reading require some attentional resources. (Consider, for example, reading an unfamiliar word or a word written in sloppy handwriting). These findings, in addition to other experimental findings described in Cohen, Dunbar and McClelland suggest that automatic and controlled processes, rather than being dichotomous, appear to lie along a continuum (see Figures 2 and 3). Both kinds of processes are assumed to require attention, but to varying degrees.

Figure 1. This graph illustrates the presumed dichotomy between automatic and controlled cognitive processes as a function of attentional requirements. Historically, it has been assumed that automatic processes require little attentional resources, and that controlled processes draw more heavily upon attentional resources, indicating that automaticity is an all-or-none phenomenon.

Figure 2. This graph illustrates the proposed continuum with automatic and controlled cognitive processes as a function of attentional requirements.

Figure 3. This figure illustrates the effect of practice on attentional demands. As you practice a skill your performance on that skill will eventually require fewer attentional resources; therefore, controlled processes are simply those with which you have had less exposure or practice.

Procedure

The materials for this experiment includes a set of 3 cards. On one card the names of four colors (red, green, blue, black) are printed in black ink. Each color name is printed 12 times, and the 48 words (4 names 12 times each) are listed in a random order. On the second card 48 patches of red, green, blue, yellow are printed in a different random order. The third card is like the first one in that it contains a list of color names, but on this card the names are printed in colored ink, and the color of the ink is always different from the color name. For example, the word "blue" might be typed in green ink, and the word "red" might be typed in yellow ink.

To conduct the experiment, the class will be divided into groups of three, and each member of each group will serve as experimenter, participant, and data recorder. Present the three cards to the participant in a random order (different orders should be used for each participant in a group). For the card with color names in black ink, the participant's task is to read the words aloud; for the card with color patches, the participant's task is to name the color of each patch; for the card with color names in colored ink, the participant's task is to name the ink color in which each word is printed. (For example, if the word "red" is printed in blue ink, the participant should say "blue".) For each card, the participant should proceed as rapidly as possible. If a mistake is made, the participant should correct the mistake before proceeding. Record the time (in seconds) that it takes for the participant to complete each card. As an extra control, you can have the participant read the words (rather than name the colors) on the card with color words in colored ink.

Results

At the appropriate time, your laboratory instructor will ask you to report your data and will provide you with summary data for the entire class.

Discussion

1. Why should the random orders of the three sheets be different?
2. Do the results for the 4 conditions differ? How would you characterize the pattern of results?
3. Explain the results in terms of the contrast between automatic and controlled processes.
4. If the stimuli for this experiment were less familiar color names (e.g., tan and maroon), do you predict that the results would have changed as well? If so, how would the results have changed, and why? If not, why not? Redraw Figure 2 by relabeling the x-axis to support your response.

Terms and Concepts

• Stroop effect
• automatic processing
• contolled processing
• selective attention

References

Stroop, J. R. (1935). Studies of inference in serial verbal reactions. Journal of Experimental Psychology, 18, 643-662.

Cohen, J. D., Dunbar, K. & McClelland, J. L. (1990). On the control of automatic processes: a parallel distributed processing account of the stroop effect. Psychological Review, 97(3), 332-361.