INTRODUCTION The Attentional Blink Experiment aims to determine the capability of an individual to recognize both the targets given that he or she is subjected to rapidly changing stimuli. Moreover, the theory states that after the detection of the first target in a rapid stream of visual stimuli, the second target is missed (Niewenstain, Potter, & Theeuwes, 2009). Hence, the experiment means to prove whether attentional blink is present in the experiment and if the theory is correct.
Furthermore, the suggested hypothesis for this experiment is that the higher separation of the two targets with each other will increase the probability of discriminating and reporting the second target with respect to the first. In addition, the experiment was conducted inside the ergonomics laboratory at the Science and Technology research building on February 5, 2013 using the Wadsworth Coglab program application. It was done at only one site to ensure the consistency of the environment. Also, each test consisted of 100 trials. I. PROBLEM STATEMENT Attentional blink is present between targets of short separation.
II. OBJECTIVES 1. Aims to confirm the presence of attention blink in the different subjects. 2. Aims to show that the percentage reported for the 2nd target increases as the separation of the two targets increases through the use of statistical analysis. 3. To identify improvements for the report of the second target in the stimulus presentation, assuming the theory is correct. III. METHODOLOGY A. Selection of Subjects The minimum required subjects was fifteen subjects which consists of the students of the present Ergcog2 laboratory class, and they were asked to answer the attention blink experiment honestly.
The group decided to add additional of 10 subjects outside of DLSU with the same conditions given to the first fifteen subjects in the class. This was done for the reason that more data leads to more consistent and less biased results. There was no particular reason nor criteria used in choosing the subjects. They were chosen out of convenience. Apparently, the subjects chosen were composed of both male and female and all subjects were in between the ages of 18-22 years old. B. Experiment Proper 1. Fifteen subjects (from the class) and ten subjects (outside DLSU) were chosen to answer the experiment on attention blink.
They were chosen using convenience sampling distribution. 2. There are two trials in this experiment and the group considered this factor. Trial 1: Subjects took the experiment without being distracted. Trial 2: Subjects took the experiment while being disturbed during the whole experimental period. Subjects were having simultaneous conversation during the whole experiment. 3. The software is activated. Pressing the spacebar indicates the start of the first trial where a sequence of letters appears. Each letter in the sequence is only flashed for 100 milliseconds. 4.
The task of the subjects is to determine if letter J, letter K, or both letters were flashed in each sequence. 5. The subject presses the “J” and “K” keys to indicate that the letters “J” and “K” were flashed in the sequence respectively. The subject can also press both “J” and “K” keys if he/she believes that both letters were flashed. 6. The keys that were pressed by the subjects are flashed immediately in the screen for the subjects to be able to check whether the software was able to receive the information correctly or not. 7. Space bar is pressed by the subject to proceed to the next trial. . After the 100 trials, a window appears which shows the graphical result of the test that was done by the subject. The graph shows the rate of how the subjects were able to detect the targets due to how the targets were separated. 9. The results were analysed and conclusions and recommendations were made at the end of the experiment. C. Tools Used * Computers with CogLab Software are used to run the trials in which data are gathered. D. Possible Causes of Error (Factors) Fatigue of the subjects is a possible cause of error in the experiment.
One run is composed of 100 trials, which can be very tiring for the eyes. As a result, the subject’s ability to detect targets may deteriorate at the latter trials of the experiment. Environmental factors can also be a possible cause of error like having noise in the background or having a conversation while doing the test. This is to test whether this kind of factor has a significant effect on the ability of the subject to detect targets. The subjects not taking the attention blink test seriously may also be a possible source of error in the experiment.
Some subjects may have just rushed the test. How the subjects would take the experiment is solely dependent on their level of seriousness. IV. RESULTS & DISCUSSION Table 1. Summary of the Mean and Std. Deviation Response on 1st target | Separation target| | 0| 2| 4| 6| 8| Mean (percent)| 56| 54. 5| 58| 54. 5| 58| Std. Deviation| 11. 7| 13. 4| 17. 2| 15. 5| 19. 6| Figure 1. Percent Response Vs Target Separation for 1st target Table 1 shows that for the first target the average responses for the 5 separation target are near to each other.
The results for each target separation might be varied for the subjects as seen in the deviations which are at the range of 11. 7 to 19. 6, but comparing the 5 mean would only result to a standard deviation of 1. 75. This means that the results are almost constant and has minimal deviation. Figure 1 also shows this trend that the % responses for each target separation are near each other. Looking at the results it can also be seen that the subjects can only see 54. 5% to 58% of the 1st target, since fluctuations in the graph is within these range. Table 2. Summary of the Mean and Std.
Deviation Response on 2nd target | Separation target| | 0| 2| 4| 6| 8| Mean (percent)| 5. 0| 39. 0| 42. 5| 58. 5| 60. 5| Std. Deviation| 6. 2| 16. 6| 11. 4| 11. 6| 15. 7| Figure 2. Percent Response Vs Target Separation for 2nd target Table 2 shows that the percent mean of the responses increases as the target separation increases. Again the results for each target separation also varied for the subjects since the deviation ranges from 6. 2 to 16. 6. But the deviation for the zero separation is not that big compared to the others, since most of the respondents here cannot detect the 2nd target.
The deviation for each target separation might be big but the data and Figure 2 would show a linear relationship with between the % response and the target separation of the 2nd target. This means that the respondents are able to detect the 2nd target more as the separation between the two target increases. The % response of the respondents for the 2nd target is from 5% to 60. 5%. Figure 3. Percent Response Vs Target Separation for 1st and 2nd target Figure 3 would show a clearer relationship between the 1st target and the 2nd target.
The line for 1st target (blue) would show an almost straight line pattern while the line for the 2nd target (red) would show a line that increases as target separation increases. The graph also shows that for target separation 0 to 4, the 1st target has a higher % response. But when the separation became 6 and 8 the 2nd target is seen more by the respondents. We could also see that the deviation between the 1st and 2nd target decreases as the target separation increases. For the 0 target separation the difference between the two targets are 51% for the 2 sec target separation it became 15. % and the difference becomes smaller as target separation increases. The best result is seen in the 8 sec target separation since 1st target has a 58% response and the 2nd is 60% response the difference between the two is only 2%. In addition, in order to identify the targets better the subject only focuses on the letters “J” and “K” and disregards the other letters in the series. In fact, this selective nature of perception would lessen the overloading of information. According to Reed (2004), selectivity is defined as the focusing of aspects of attention, wherein the subject pays attention to some aspects while ignores the others.
To prove that the attentional blink theory is correct in stating that the first target is unaffected by the separation of the signals. And the second target, on the other hand, shows that the longer the separation period of the first signal to the second, the higher the response (Mackewn & Goldthwaithe, 2004). Regression technique is used to see the relationship between the target separation and % response of the 1st and 2nd target. This would show how the target separation (independent) affects the detection of the target for the 1st and 2nd target (dependent). Table 3.
Regression summary for 1st target. N= 50| Beta| Std. Err. of Beta| B| Std. Err. of B| t(48)| p-level| Intercept| | | 55. 4| 3. 75| 14. 77| 0. 00| Separation| 0. 04| 0. 14| 0. 20| 0. 77| 0. 26| 0. 80| The regression summary would show that the separation of the target is not related with the percent response of the 1st target since the p-level of the regression is 0. 80 meaning it is not significant in identifying the value for the 1st target. Table 4. Regression summary for 2nd target. N= 50| Beta| Std. Err. of Beta| B| Std. Err. of B| t(48)| p-level| Intercept| | | 15| 3. 9| 4. 18| 0. 00| Separation| 0. 79| 0. 09| 6. 53| 0. 73| 8. 92| 0. 00| The regression summary shows that for the 2nd target the target separation is significant since a p-level of 0. 00 is shown. Therefore, this means that target separation affects the % response for the 2nd target. On the other hand the beta value of 6. 53 shows that as the target separation increases the % response for the 2nd target also increases. The Attentional Capacity Theory Duncan et al. have proposed that T1 occupies attentional capacity to the detriment of a trailing T2 target.
This theory suggests that the duration for which T1 continues to occupy attentional capacity is related directly to the T2 processing difficult (Rochester Institute of Technology). This explains why the 2nd target increases as the separation time increases. It is because the theory states that every person has their own attentional capacity and if separation time is bigger the information processing do not overlap and the two targets are seen by the respondent. This also why the first letter is first seen since it is the one that occupies the person’s attentional capacity first and is first processed by the person.
Outside Factors In the conduction of the experiment, although the distraction may have a small effect on the signal detection of the subjects, the results as shown in the graphs make it clear that attentional blink is not affected by the amount of external distraction since it is an internal issue. As mentioned, in trial 2 the subjects were distracted by assigning someone close enough to generate a conversation with them through the whole duration of the experiment. There is no significant difference found in the detection of the targets between being distracted and not.
This is because the subjects were observed to say “ha? ” more often than not during the conversation. Having their attention focused on the experiment applies the theory of selective attention wherein one tries to pay attention to one input in the presence of others (Glass & Holyoak, 2004). Visual dominance is another concept that can be seen in the experiment. It can be observed that visual targets dominate over auditory targets (Glass & Holyoak, 2004). This also explains why the subject is more inclined in doing the experiment rather than chatting with the distracter. V. CONCLUSION & RECOMMENDATIONS
Based on the results of the experiment, it can be concluded that the theory of attentional blink is correct. The hypothesis made at the beginning could be verified by the results obtained. These results show that separation does not have a large effect on the probability that the first target would not be detected by the respondents, since the average percentage reported for the first target by the respondent is relatively unaffected by separation. The values were close to each other. On the other hand, average percentage reported for the second target increased as the separation increased.
This shows that the hypothesis that the longer the separation between the targets, the higher the chance of getting the targets right is correct. There are less chances of an attentional blink when more letters are in between, since the second letter is delayed. This gives the user a chance to have more accurate results. Although the program was effective in proving that the Attentional Blink Theory is correct, there could still be some improvements that could be done. Since the program has a black background and white letters for the stimuli, a way of making the second target easier to detect is to add color or change the background color.
According to (Pashler, 1999), the second target could be easier to detect when there is color discrimination. When provided, it seems to cause the blink to virtually disappear because there is a different dimension. Sizes of the letters were the same for all. Biased attention may occur when the contrast and sizes of the targets differ (Proulx & Egeth, 2006). In the research conducted by Proulx and Egeth (2006), objects with better luminance contrast are processed rapidly and precisely compared to lower contrast items. It also shows that larger objects can influence visual performance.
In order for the subject to identify the second target, the size of the signals or targets can be made bigger. A sample for this is illustrated below. Based from observation, the respondents made mistakes on entering what letter they saw. The program did not allow the respondent to change his or her answer. A recommendation for the enhancement of the program could be having the function to let the respondent change his or her answer, so that the respondents’ probability of getting the correct answer would increase. This in turn can improve the respondents’ data.