Dse212Tma_03

An Adapted Stroop Experiment: Investigation of interference between automatic and controlled processes by measuring response time using colour-related and colour-neutral words. Abstract The interference effect between automatic and controlled processing tasks was examined using an adapted Stroop experiment. Past work showed that response times in naming ink colours of printed words was longer when the words were colour names. This result was interpreted as an example of interference between automatic (reading) and controlled (colour naming) processing tasks. In this experiment, instead of colour names, colour-related words were used to see if the interference effect would still occur. Response times were greater when the words came from a list of colour-related words rather than one of colour-neutral words. This agreed with the findings of the previous experiment indicating that interference had occurred. Introduction Our senses receive copious information from many stimuli, which undergoes internal cognitive processes. All this information cannot be processed simultaneously. A cognitive process called attention selects information from what is sensed, enabling it to be perceived and processed further, other information is disregarded, i.e. not attended to (DSE212 Course Team, 2007). As the brain has limited capacity and resources, it is expected that only some of the available information will be taken in and processed. Simon and Levin (as cited in Edgar, 2007) designed a study to test this theory. Their study showed that most people failed to detect visual changes, referred to as ‘change blindness’ (DSE212 Course Team, 2007). One explanation for why we cannot attend to everything, suggested by Kahneman (as cited in Edgar, 2007) is that the brain has a ‘limited capacity central processor’ to process sensory information and combine it with stored information. According to Kahneman (as cited in Edgar, 2007), there is one main pool of resources to filter information with limited capacity. This limit varies with arousal level. If arousal level is higher, more information can be taken in and processed. There are two types of attentional information processing. One type is consciously handled, and is referred to as controlled processing. Controlled processing makes heavy demands on attentional resources, takes time, engages conscious awareness and is affected by limited capacity (Edgar, 2007). The studies of Schneider and Shiffrin (as cited in Edgar, 2007) suggested that there are also automatic processes, which make little or no demand on attentional resources, are faster, require no conscious awareness and are unaffected by limited capacity. Automatic processing has advantages. As it is independent of conscious processing, there will be more resources to process other tasks. Some experiments for dual tasks Mcleod (as cited in Edgar, 2007) showed that two or more tasks can be performed at the same time if information is drawn from separate pools. However, Kahneman (as cited in Edgar, 2007) underlined disadvantages of doing multiple tasks simultaneously. Tasks drawing information from a common pool of resources interfere with each other (Edgar, 2007). For instance, reading and talking cannot be performed simultaneously. Posner and Boies’s (as cited in Edgar, 2007) studies supported showing that response times declined when participants were required to do two things simultaneously. The disadvantage of automatic processing is that it interferes with controlled processes, because it is impossible to stop them voluntarily. For instance automatic processes such as reading interferes with tasks requiring controlled processing e.g. naming a word’s ink colour. This is demonstrated by a phenomena referred as the Stroop effect, which shows the interaction between automatic and controlled processing (Edgar, 2007). Stroop (as cited in Edgar, 2007) conducted a study revealing that if the name of a colour is printed in ink of a different colour, for example ‘red’ printed in blue ink, naming the colour of the word is more difficult, indicated by the fact that response time is longer than if the word is a colour-neutral word. This work represented an extension of the research described above aiming to determine whether interference takes place between the automatic process of reading and the controlled process of colour naming of words drawn from two lists in an adapted Stroop test. Here, colour-related words, instead of colour name words were used. The hypothesis was that response time for naming the ink of colour related words will be longer than for naming ink colours of neutral words. This is a one-tailed hypothesis. Null hypothesis was that there is no difference in response time for two word types. Method Design A within-participant design was used. The independent variable consisted of two conditions. Firstly, a colour-related word list (experimental condition/stroop condition) in which the words were printed in a colour different to that associated with the word. The second condition was a list of colour-neutral words (control condition) in which the words were similar to the colour-related words and printed in the same colours. Each word appeared five times at random. Participants had to name the colour of the ink in each condition. The dependent variable was response time to name the ink colour of all the words on the list. It was measured in seconds by the researcher using a digital timer. The same instructions were read to each participant. Participants were numbered from 1 to 20. Odd numbered participants carried out condition 1 first and then condition 2. Even numbered participants performed the task in the reverse order. This was to counterbalance for order effects. Participants 20 participants took part in this experiment. 16 of the participants were recruited by Open University and the remaining 4 were recruited by the researcher from family members and friends who volunteered with no incentive offer. All participants were asked to wear any glasses they normally use for reading. All participants stated they had normal colour vision. They were all native English speakers and able to read. The age of the participants varied from 30 to 70 years. There were 10 females and 10 males. All participants were naïve to the nature of the experiment, however, they were briefed in advance, their informed consent was gained, their right of withdrawal at any stage was explained and they were debriefed in the end. Materials The stimuli shown to the participants were lists containing 30 words in two columns on white paper. The six colour-related words, BLOOD, LEMON, GRASS, CARROT, PLUM, SKY, were chosen as the experimental (stroop) condition. The ink colour of those words did not match with the words. For example, LEMON was printed in colours other than yellow. The control condition used the six colour-neutral words of BLAME, LEDGE, GRADE, CAREER, PLAN, STY printed in the same range of colours as the experimental condition. These were deliberately chosen to contain similar letters and have the same length. In each condition, all words printed five times in a random sequence giving total of 30 words. A copy of each condition list is provided in Appendix 1. Standardised written instructions were given to all participants (see Appendix 2) and a copy of the consent form was read and signed by all participants (see Appendix 3). A data sheet was prepared to record the data for all participants (see Appendix 4) and a digital timer was used to measure response time, i.e. how long it took each participant to complete each condition. Procedure The participants were asked if they would like to take part individually in a psychology experiment that was related to naming ink colour of two lists of words. Participants were checked for their ability of reading and speaking English, recognising colours and having normal vision as well as asked whether they were familiar with the Stroop effect. When these had been done and the participants had agreed to take part in the research, their written consent was obtained (see Appendix 3). Participant’s age and sex was recorded. The experimental procedure was explained by reading the standardised instructions (Appendix 2) and an example was given. These explained to the participants that they shown two lists of words and that they should work through them naming the ink colour of each word aloud. They were instructed to go through both lists as quickly as they could. When it had been confirmed the participant understood the task, the two lists of words were placed faced down on a table. When the participant was ready, the first list was turned over and the digital timer started. When task was completed, the timer was stopped and response time was recorded to the nearest second on the data sheet. The second list was then immediately placed on the table and turned over and the same procedure was duplicated. Odd numbered participants began with Condition 1 before going onto Condition 2. Even numbered participants did the reverse. Completion of the experiment was followed by debriefing of the participants. This consisted of a brief explanation of the Stroop effect and the differences between the two lists. The participants were encouraged to ask any questions and they were assured that their anonymity would be respected. Finally they were thanked for taking place in this experiment. Results The research hypothesis was that response time for naming the ink of colour related words will be longer than for naming ink colours of neutral words. Response time taken by each participant to complete each condition was measured to the nearest seconds and recoded (see Appendix 4). Table 1. Mean and standard deviation for response time output in seconds Condition Mean Response Time (in seconds) Standard Deviation Colour-related words (Condition 1) 26.65 4.934 Colour- neutral words (Condition 2) 24.35 3.842 Table 1 indicates the mean response times and standard deviations were different for both conditions. The colour related words (condition 1) had a longer average response by 2.30seconds compared to the colour- neutral words (condition 2). The standard deviation shows that response time dispersion was greater in condition 1 than condition 2. The data was analysed by using a paired-samples t-test, which showed there was statistically significant difference between the participants mean response times for each condition (t(19)= 3.397, p=.0015,one-tailed, d=.524 and this was not due to sampling error (p value smaller than .05) (see Appendix 5). Size effect shows a difference of 0.5 standard deviation points, which can be considered medium effect size. According to this result, the hypothesis was accepted and null hypothesis was rejected. Discussion The results supported the research hypothesis. They reveal that the response time was significantly longer when the task involved naming the ink colour of words taken from the colour-related word list compared to when they were taken from the colour neutral word list. This is consistent with the Stroop effect and is in accordance with Kahneman’s idea that if multiple tasks need to draw on a single pool of resources, interference occurs. The results also demonstrate the interaction between automatic and controlled processes, described by Schneider and Shiffrins (as cited in Edgar, 2007). Firstly automatic processing needs to be suppressed to allow naming of the ink colour of the colour-related words rather than the colour associated with the colour-related word (e.g. green for grass). To do this, controlled processing must be engaged to name the correct ink colour. This requires longer processing time, as controlled processes require conscious effort, unlike automatic processes which use less resources and effort, thereby consuming less time. It follows that reading (automatic process) is processed before the colour naming (controlled) response is made. When mean response times are compared between even and odd numbered participants on a bar chart, it is seen that there is a smaller difference in mean response times for the two conditions when condition 2 is completed first (see Appendix 6). When the participants (even numbered) completed the neutral word list (condition 2) first rather than colour-related word list, they may have found a technique to help the colour naming task such as concentrating on the first letter of words instead of reading the whole word (Edgar, 2007). This avoids the automatic processing needed to read a word and avoids the colour association interference induced by reading from the colour related word list (condition 1), giving a quicker response time by minimising interference. This illustrates the need for counterbalancing to minimise order effects in the experiment. The experiment takes no account of how strongly a participant associates a colour with a word. Association may be stronger for certain words, this may vary between individuals. Thus, the degree of interference caused by reading from the colour related word list may vary between individuals confounding results. Other shortcomings of the experiment include the relatively small sample size. Also the task involved in the experiments was not one typically performed in everyday life and so ecological validity was lacking. No account was taken of whether responses were accurate in condition 1, i.e. whether a participant named the ink colour used in a word correctly or confused it with the word’s associated colour. Future experiments could investigate factors affecting the number of conscious errors and unconscious errors made. Obviously, the inability of controlled attention to spot mistakes could have serious implications. Whilst the experiment agrees with results of the experiment conducted on colour name words which also showed the Stroop effect it would be interesting to quantify the relative size of the Stroop effects for the case of colour name and colour related words. This could be done with a 3 condition experiment using lists of colour -neutral, colour- related, and colour name words. It might be expected that the Stroop effect would be larger for the colour name words than the colour related words as the interference effect should be greater. (2150 words) References Edgar, G. (2007). Attention and Perception. In D. Miell, A. Phoenix, & K. Thomas (Eds.), Mapping Psychology (2nd ed., pp. 1-56). Milton Keynes: The Open University. Appendices Appendix 1: Stimuli (Word List) Appendix 2: The Instructions Appendix 3: Consent Form Appendix 4: Raw Data Appendix 5: SPSS Output for Inferential Statistics Appendix 6: SPSS Output for Bar Chart APPENDIX 1: Stimuli (Word List) Condition 1 SKY PLUM PLUM BLOOD LEMON LEMON GRASS GRASS CARROT BLOOD BLOOD SKY PLUM CARROT CARROT LEMON SKY PLUM GRASS GRASS BLOOD CARROT LEMON SKY CARROT BLOOD GRASS LEMON SKY PLUM Condition 2 STY PLAN PLAN BLAME LEDGE LEDGE GRADE GRADE CAREER BLAME BLAME STY PLAN CAREER CAREER LEDGE STY PLAN GRADE GRADE BLAME CAREER LEDGE STY CAREER BLAME GRADE LEDGE STY PLAN APPENDIX 2: The Instructions In a moment I will place a sheet of A4 paper in front of you that contains two columns of words. You will notice that the words are written in six different colours of ink – red, blue, green, yellow, orange and purple. What I would like you to do is to say out loud the colour of the ink each word is written in. Start with the word at the top of the left column and work downwards. When you have finished all the words in the left column, start on the right column. Remember, I do not want you to read the word itself out to me, instead I want you to state what colour the ink has been written in. You should work through the list as quickly as you can. To help you, here are two examples: CHAIR* For the item above you would respond ‘blue’. HOUSE* For the item above you would respond ‘red’. Do you understand what you will be required to do' (If yes, then proceed to task. If no, go through the examples again) APPENDIX 3: Consent Form Consent form I have been asked to participate in an experiment that investigates one aspect of cognitive psychology and give my free consent by signing this form.  I have been informed about the research and why it is taking place.  I understand that my participation in this research is voluntary.  I understand that I can withdraw from the research at any time.  I understand that my data will be anonymous.  I understand that I will be provided with a debrief after taking part in the experiment. Signature _______________________________ Date ____________________________________ APPENDIX 4: Raw Data Table 2 - Data from sixteen participants in the colour-associated words experiment Participant number Age (years) Sex (male/female) Condition 1 (colour-related words) time (seconds) Condition 2 (neutral words) time (seconds) 1 48 male 22.00 23.00 2 43 male 31.00 26.00 3 38 female 26.00 25.00 4 31 male 20.00 23.00 5 18 female 21.00 18.00 6 47 female 36.00 28.00 7 39 male 21.00 16.00 8 22 male 28.00 24.00 9 37 female 30.00 28.00 10 42 female 18.00 21.00 11 55 female 22.00 19.00 12 68 male 25.00 22.00 13 20 female 32.00 25.00 14 69 male 29.00 27.00 15 39 female 32.00 30.00 16 42 female 26.00 24.00 17 32 female 25.00 23.00 18 44 male 26.00 29.00 19 42 male 31.00 27.00 20 54 male 32.00 29.00 Notes: 1- Data on odd-numbered rows (1, 3 etc) are from participants who did Condition 1 then Condition 2, whilst data on even-numbered rows (2, 4 etc.) are from participants who did Condition 2 then Condition 1. 2- The data collected were measured to the nearest second. APPENDIX 5: SPPS OUTPUT FOR INFERENTIAL STATISTICS Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 colour-related response time 26.6500 20 4.93404 1.10328 colour-neutral response time 24.3500 20 3.84263 .85924 Paired Samples Correlations N Correlation Sig. Pair 1 colour-related response time & colour-neutral response time 20 .790 .000 Paired Samples Test Paired Differences t df Sig. (2-tailed) Mean Std. Deviation Std. Error Mean 95% Confidence Interval of the Difference Lower Upper Pair 1 colour -related response time- colour-neutral response time 2.30000 3.02794 .67707 .88288 3.71712 3.397 19 .003 Note: p = .003 was halved as the hypothesis was one-tailed. Therefore the actual result for consideration was p = .0015 APPENDIX 6: BAR CHART Mean Response Time When Counterbalancing Odd and Even Numbered Participants