Hypotheses
Research Aim
The aim of the study is to discover if automatic activities (EG reading) can interfere with other (controlled attention) tasks (EG correctly identifying colours).
Experimental Hypotheses
There will be a significantly longer average time taken to finish reading the list of words written in conflicting colours, compared to reading the list of words written in colours, which match the word.
Null Hypotheses
There will be no significant difference in the time taken to read the two word lists, and any difference that does occur is due to chance alone.
Independent Variable
If the word list is conflicting or normal.
Dependent Variable
The difference in time taken to read the two word lists.
Methodology
Experimental Method
Laboratory experiment. This was chosen because it allows for the best control of extraneous variables.
Experimental Design
Repeated measures design was chosen, as it was most appropriate for this experiment. The advantage of this design is that it avoids the unreliability that individual differences could bring to the results, as one participant is used for both conditions. Order effects however, more specifically improvement in performance due to practice, were a problem with this design. To combat this half the participants were given the lists in reverse order with the conflicting list first, and the same list second.
Sampling and Sampling Method
10 participants were used, (5 male and 5 female), aged 16-19. The sample was obtained in the college library, and was an opportunity sample. I chose this sampling method because it was quick and easy to obtain, although not truly representative of the population. The participants all had normal colour vision and were not dyslexic.
Materials
- 3 normal/same lists of words
- 3 conflicting lists of works
- Stopwatch
To compile the normal word lists I created 3 pages (20 words per page written in 4 colours) all in the same font type and size, on a word-processing package. The words were randomly listed on the page (I made sure there were no patterns in the word list by reading them) and were the same colour as the word that was written. I followed this method for the 3 conflicting word lists, except changed the colour of the word so it was different from the written word on the page. See appendix 2 for word lists.
Procedure
Participants were read the standardised instructions (see appendix 1), after which the 6 word lists were placed face down on the table. The lists (see appendix 2) were arranged with the conflicting lists first for 5 participants and the normal lists first for the other 5 participants in order to combat order effects. When told to ‘go’, the stopwatch was started and the participant turned over the 1st word list. When he/she finished reading they said ‘stop’, and the stopwatch was stopped. The time was then recorded, and this was repeated until all 6-word list times were recorded. The participants were then de-briefed (see appendix 3).
Controls
The materials used, standardised instructions and de-briefing were kept the same for all participants. Also, seeing as the whole experiment was conducted in the library the level of background noise, and other extraneous variables were reasonably controlled. All participants saw the word lists for the same time, only turning them over when the experimenter said ‘go’. Counter-balancing controlled order effects. The lists were left face down on the table when the participant was not reading them to ensure they could not see the list, and so ‘cheat’ before it was time to read that particular list. 6 lists were used to achieve a more reliable set of results.
Ethics
All participants were over the age of 16, and had given informed consent. They were told they had the right to withdraw at any time, and were thoroughly debriefed. Nobody was put under any undue stress during the experiment, although some participants may have felt slight frustration when attempting the Stroop task. No embarrassment was caused to any of the participants.
Results
As the table below shows, it took significantly longer for participants to read out the Stroop or conflicting list of words, than it does to read the normal list of words.
Table of Average Time Taken to Read Each Single List
The range of the Non-conflicting list of words was 5.88 seconds, and the range of the Stroop list of words was 12.52 seconds. The modes of each condition can be seen in the histograms in appendix 5. As can be seen, for the non-conflicting list of words the modal number is 9, and for the conflicting list, the modal number is 14.
It took on average, 6.9 seconds longer for each participant to read the Stroop lists of words, showing that interference has occurred. The raw data can be seen in appendix 4.
Statistical Analysis
I used a related or correlated t-test to analyse the raw data because the design was related, I predicted a difference in the results and it was interval/ratio data I was using. I used the data for a two-tailed test, as I had predicted both difference and direction in my hypotheses.
The analysis gave a result of t = 2.9547004 or t = 2.96. The degrees of freedom (df) were 9, and the results were significant to the p<0.01 level. This means there is less than a 1% chance these results are due to chance alone. I accept the experimental hypotheses, and reject the null hypotheses at the 0.01% significance level. The calculations for this test can be seen in appendix 6.
Discussion
The results supported the experimental hypothesis, and demonstrated there is a significant difference in the time taken to read each list due to the interference of the automatic process, in this case reading.
Although the sample size was small, and did not truly represent the population, the results are clear that the Stroop effect does exist. This evidence also supports the theories mentioned in the background information, reading, an automatic activity, hinders the participants performance in the task of reading the colours of the words on the paper.
I controlled all the extraneous variables, the briefing, the standardised instructions, the de-briefing, and time and order the participants saw the lists in. There were no confounding variables in my experiment, however, the experimental design was repeated measures, and so unavoidably there could be a problem with order effects, (where a participants improves at a task due to practise). The experiment was performed in a library; therefore background noise could have affected their performance, although this was equal for all participants. The sample was in no way androcentric, (gender biased), and although the age range was small, I believe this did not affect the results in any way.
The experiment did lack ecological validity, as the participants are unlikely to come across a list of words like this in their normal life. However, this could be combated by asking them to read a story or paragraph with the Stroop effect words in it, and recording if they take longer to name the non-conflicting colours used in the paragraph, or if they take longer to name the Stroop effect colours used in the second paragraph.
However, as mentioned by Logan (1980), controlled processes can become automatic if used repeatedly, therefore if asked to perform this task many times, it could become automatic. Therefore, instead of being under conscious control it would become unconscious like normal reading and require less time and concentration.
If I was to repeat the experiment again, there are other factors I would like to investigate, to determine whether they have an effect on the results or not.
These are:
- The experimental design. Using matched pairs (matched on reading speed) would control all order effects.
- A different sampling method. Using a stratified sample would allow the results to be more representative of the true population.
Further Research Ideas
- Children. Trying this experiment with a small child who has not yet learned to read would provide more data to support the theory, because reading has not yet become automatic for them.
References
Logan G. D. (1985). Attention and automacity in Stroop and priming tasks
Taken from:
Shiffrin R. M & Schneider, W (1977) Controlled and automatic information processing
Taken from:
Stroop J. R. (1935) Studies of interference in serial-verbal reaction
Richard Gross 2001
“Psychology, the science of mind and behaviour”
Hodder & Stoughton – London Page 193-194