Based on expectations affecting our perception of visual illusions, it was expected that people given prompts would be more likely to see both forms of the visual illusions, and so the hypothesis was developed around this. The rationale of this experiment was to see if perceptual set exists amongst students.
Hypotheses
Research hypothesis: There will be a significant difference in the number of illusory figures seen by a group if participants given prompts than those not when presented with ambiguous figures.
Null hypothesis: There will be no significant difference in the results of the two conditions.
Methodology
Method/Design
The experimental method for this study is a laboratory experiment; this method was chosen as control over extraneous variables was necessary. This method suffers from demand characteristics, but this was overcome by a control. The experimental design of this study was independent measures. The advantages of this design are that the same test can be used (which is essential for this study), and it reduces order effects. The independent variable of this experiment was whether a prompt was given to the participant in the question or not, the dependant variable was the score out of ten that the participants achieve.
Sample
The study used an opportunity sample, in which the population was made up of students, and the sampling frame was students from a sixth form college aged between 16 and 18. Altogether, forty participants were tested in this study, twenty randomly assigned to each condition by tossing a coin to avoid order effects.
Procedure
The images that were shown to the participants were all images which could be seen in two forms, so that they could say whether they could see both forms of the image or not and be scored accordingly. There were ten images in total; shown one after another to the participants (see script in appendix). When the participants were being shown the images they were asked either “How many things do you see?” if in the prompt condition, and otherwise “What do you see?” if in the non-prompt condition. Each participant was tested individually and shown all ten images. This is a standardised procedure which can easily be replicated.
Controls
If participants said that they could see both forms of the image, they were asked to outline both forms of the image, in order to ensure that they could actually see both forms of the image and that they were not lying due to demand characteristics.
Ethics
The participants gave consent, but were not told the aim of the experiment, this was necessary to reduce demand characteristics. There was no personal data gathered, participants had the right to withdraw at any time, the participants did not suffer any harm, and at the end of the experiment all participants were debriefed.
Measurement and analysis of data
The participants were given a score of either 1 if they could see both forms of the image or 0 if they couldn’t for each image. After each test on a participant, they were given a score out of ten for all of the images. These scores were used in the statistical tests.
Results
From the raw data, the mean, median and mode were calculated:
A graph was also made to represent the results:
The graph and results show that the scores were much higher in the prompt condition than in the non prompt condition.
A Mann-Whitney U test was carried out:
From the results of this test it can be seen that U1 is significantly lower than the critical value, and that U2 is significantly higher than the critical value, it can therefore be said that the results are significant at the p= 0.05 level.
This test shows statistical evidence that if people are given prompts, they are more likely to see both forms of visual illusions.
The standard deviation was also calculated, the results of which are as follows:
The standard deviation shows that there was very little spread in the data.
Discussion
The results were statistically significant as shown by the Mann-Whitney U test to be significant at the p=0.05 level, this means that the original experimental hypothesis was correct as so the null hypothesis can therefore be rejected. It can therefore be said that giving a person an expectation affects their perception of a visual illusion, and perceptual set exists in students. As the method was a standardised procedure, it could easily be replicated and get the same results as it is not era dependant, therefore the results have high reliability. However, as all the images could be seen in two different ways, the participants could have known from previous images that there is meant to be two forms of the image, and so may have actually been given an indirect prompt, and this may have accounted for some of the high results in the non prompt condition. Individual differences may also have accounted for some of the high results in the non prompt condition, but since the experiment used many different participants, this would not have had a significant effect on the final results. The participants all gave consent, had the right to withdraw, did not suffer harm, were never deceived and were all fully debriefed.
If the experiment was repeated, it should include normal images that could only be seen in one form with other images which could be seen in two forms, so that participants are not given an indirect prompt. If the participant was shown normal images together with visual illusions, then they may not necessarily always be looking for two forms of an image, and so the participants may not always be looking for two forms of the image, thus removing the indirect prompt.
The sampling technique should also be changed to a stratified sample and increase the sample size. It should also include people of different ages and different professions.
The advantages of these changes would be an increase in validity whilst keeping the procedure standardised, as well as increased ability to generalise the results to the population.
Further research could be done by using images which have more than two different forms (see below), and testing whether people can see more than two different forms if given a prompt.
This study is useful as it gives statistical evidence which supports the idea of perceptual set. Further research would be useful for air traffic control, as it is very useful for workers to be able to see more than one form of an image when looking at incoming planes, weather, etc.
References
Bugelski and Alampay:
Images:
Testing:
Further research:
Appendix
Here is the raw data collected from the experiment, together with the calculations for the statistical tests:
Calculations for Mann-Whitney-U test:
Calculations for standard deviation:
On the next page is the test used in the experiment.