Testing Reaction Times of Local and Global Perception Free Essay Example

This study examined visual perception and the rates at which global and local features are reacted to with an aim of replicating and validating a previous experiment conducted by Navon (1977) to see if global processing was faster than local processing.

There was 222 University of Newcastle students participating in the experiment, partaking in two phases, one centred round global processing, the other around local processing, where there reaction times were recorded using a computer program and imputed into a data worksheet.

Results indicated that, as predicted, global processing occurred at a faster rate than local processing. It was concluded that global features were processed at a faster rate than local features, giving evidence that more attention and focus is needed to identify the more intricate features of an object, however future studies should improve the validity of their studies by employing better randomisation in the gathering of participants.

Visual Perception is an efficient and flexible process( Lin, Lin, & Han, 2008) within the eye that assists humans and animals in acquiring information about their settings by detecting light that is reflected from surfaces, allowing individuals to understand what objects are present and the appropriate behaviour to respond in (Yantis, 2001).

Understanding perception and the types of processes that are involved is vital in determining whether we perceive a scene feature by feature or whether the process is immediate upon the visualisation of the object (Navon, 1977).

The methods involved in the processing of visual input selects the information that is worth receiving and attending to and focuses more attention on recognising the objects and features of that input.

Many studies have focused their research on the hierarchical levels of perception as the focus of the manner in which visual processing is carried out, were larger features are constructed by an appropriate arrangement of minor features (Kimchi, Amishav, & Sulitzeanu-Kenan, 2009).

This idea suggests that we focus our immediate attention on the whole object (global), while we spend more effort and time in recognising the more detailed features of those objects (local). The whole object is known as global features, which are predominantly word bound and are what is commonly referred to as the general word shape (Schiepers, 1978) whereas local features are the individual structures of those words and are small in shape.

In relation to this, research has found that during the visual perception process of adhering to global and local tasks, our perception can be markedly modified or affected by not only our state of mind at the attention we give to that particular task, but also by the affective assets of the stimuli involved (Lin et al, 2008). In his study on visual perception and global processing, Navon (1977) reported that when an individual visualises an object, the global features of that particular object will be captured and recognised before the local features.

He puts forward the motion that people’s visual process is organized so that they progress from global structuring towards more fined tuned processing. The study also found that when global letters are comprised of identical local letters, known as congruent, the time taken to recognise the local features was less than the time taken to recognise global features comprised on different smaller letters, known as conflicting.

Participants of both Navon’s experiment and those of a study conducted by Kimchi, Amishav and Sulitzeanu- Kenan (2009), were reported to have reacted at a faster rate at a global level, as well as reporting that inconsistent information flowing from global to local process produced an interference with local level responses, but had little to no effect of global responses. In light of this information, the aim of this study was to replicate the study conducted by Navon on global and local perception, to test their theories of global precedence in visual perception and determine if they were valid.

It was therefore hypothesised that the reaction times for global judgments would be faster than the reaction times of local judgments. It was also hypothesised that consistent stimuli would be faster than conflicting stimuli in the local tasks. Method Participants The sample for this study consisted of 222 participants who were second year psychology students from the University of Newcastle. All students were participating as part of a course requirement and all had given their consent to participating in the study.

Design. The independent variables consisted of 3 consistency types (contingent, neutral and conflicting displays). The dependant variable was the reaction time of participants. Apparatus A computerised program SuperLab Pro and a keyboard was required in order to record reaction times. Stimuli In the first phase of this study that was based around recognising global features, the letters used were large H’s or S’s. In the second phase, global letters of H, S and O were used and were made up of smaller H’s or S’s, the local characters.

These make up the 3 stimulus types: congruent, neutral and conflicting. The large and small letters could be combined to create 3 types of displays: 1) Consistent, where the large and small letters were the same (H printed from small H’s, or S made of small S’s). 2) Conflicting (H made of small S’s, S made of small H’s), and 3) Neutral (H made of O’s, S made of O’s). A mask, a shaded black square, was used to avoid images lasting in our memory. Figure 1. Set of stimuli in the test phases of global and local tasks.

Figure 2. The three display types used in identifying local features in phase two. Both figure 1 and 2 were obtained from the original Navon report (Navon, 1977). Procedure Students began the experiment by adjusting the position of their head from the monitor, approximately 1 arm’s length away from the screen. They first participated in 10 practice trials for global conditions before beginning the first task. In this first task they were shown large letters made up of smaller letters. Each letter stayed on the screen for 100 milliseconds followed by a mask.

When the mask appeared, students were asked to identify whether the large letter they were shown was in the shape of a H or an S. if the letter in question was a H, students pressed the z key on the keyboard, but if it was an S they saw, they were asked to press the / key. This was repeated 36 times for global test trials and response times were recorded. They then entered the task focused on local. They again participated in 10 practiced trials for local conditions. The second test trial consisted of 36 test trials in which participants were shown large letters made up of smaller letters.

Each letter stayed on the screen briefly followed by a mask. When the mask appeared, it was the participant’s job to say whether the smaller letters that made up the larger letters were small H’s or S’s. Again, if they were H’s the z key was pressed, while if they were S’s the / key was pressed. Results All errors were first screened out from the data along with all response times below 150 milliseconds and above 200 milliseconds. Paired sample t tests were used to compare global and local tendencies. An ANOVA was also utilised in order to detect if the differences found were statistically significant.

To examine whether or not reaction times for global processing as faster than local processing, we compared the mean reaction time scores for global and local processing along the 3 types of stimuli. It was found that global processing was indeed faster than local processing, as can be seen in figure 3. The mean response times for consistent (M=400. 87, SD=139. 55), neutral (M= 406. 54, SD=145. 50) and conflicting (M=392. 39, SD= 159. 63) within global processing tasks were calculated to be responded to quicker than the mean response times of consistent (M=455. 23, SD=153. 33), neutral (M= 473. 35, SD=123. 7) and conflicting (M=504. 86, SD= 134. 21) or local tasks, as indicated in figure three, where the black line shows an increase in reaction time.

The reaction times that were recorded for global tasks were almost equally as fast across the conflicting, congruent and neutral condition type, which is demonstrated in figure 3 by the broken line. In order to calculate whether differences found were statistically significant within subject’s effects, an ANOVA was conducted on the results. The main effect of conditions, which was comprised of global and local processing, were found to be statistically significant, F (1, 221) = 93. 7, P<. 001. The ANOVA also produced a statistically significant result when comparing the main effect of consistencies (the three levels of consistency: neutral, conflicting and congruent, and there averages), F (2, 442) = 11. 14, P<. 001, indicating that … Finally, it was found that the interaction between both consistency and condition of the tasks was statistically significant, F (2, 442) = 19. 42, P<. 001. After comparing the means of global and local processing, four T tests were calculated in order to follow up the interactions and compare the global and local tendencies.

The first paired t test between global consistent and global neutral indicated that there was no significant difference in reaction time between the two tendencies, t(221) = -1. 001, p= . 318. The second paired t test focused on global consistent and conflicting tendencies with results again not being statistically significant, t (221) = 1. 265, p= . 207. These results are evident in figure 3, which shows no considerable variation in reaction times between the three of global processing, shown by the relatively flat broken line.

There was, however, statistically significant result recorded for comparisons between local consistent and neutral stimuli, and local neutral and conflicting stimuli; t (221) = -2. 866, p=. 005 and t (221) = -5. 318, p= . 001 respectively. As predicted in the second half of the hypothesis, this demonstrates that there is a difference in reaction times between the three levels of consistency and that indicates that consistent reaction time is faster than conflicting reaction time, which is represented in figure 3.

Figure 3. The mean reaction times for global and local conditions across the three consistency types Discussion The results of this experiment found that processing of global features of objects is performed at a faster rate than local features, with their also being a difference in the speed in which local features are processed across three consistency displays. These main findings of the study support the hypothesis that the processing of global judgments is faster that the processing of local judgments.

Furthermore, the results of the study partially supported our second hypothesis that the processing of consistent stimuli would be faster than the processing of conflicting stimuli in the local processing phase. The observations that have been made that the processing of global features is faster and requires less focus and attention than the processing of local features is consistent with the results and findings reported by Navon (1977) in his study on visual perception. In the results of this current study, global was processed faster than local, and within local processing producing statistically ignificant reaction times depending on the type of consistency stimuli. These results were consistent with those found by Navon and supported the notion that while most individuals can simply attend to the global levels of processing, they cannot successfully skip global processing and go straight to local processing. These results indicate that global processing is a vital step in the hierarchical stages of perception and is need before delving into more detailed analyses of smaller local features.

While this study does provide support for Navon’s global processing study, it does have certain limitations that should be considered and readjusted when carrying out future research. One limitation is the sample size. Even though 222 participants seem like quite a large sample size, it is not a very well randomised sample in the sense that all participants were academic students. Perhaps in future studies, applying a more randomised sample will produce a more accurate and scientifically reputable result.

Future studies would also benefit from increasing the number of tests conducted, a move that may increase the reliability of the results. There was found to be no obvious systematic effects on the results that were observed, with time of testing and other surrounding conditions having no significant effect on the recorded results. In conclusion, this study has revealed that global processing is an essential stage in visual perception that is required to be met before we can engage in more fine-tuned analysis on the detailed features of an object.

The results indicate that individuals react to global features at a faster rate than local features, indicating that more focused attention is needed during local processing in order to capture and interpret the appropriate information. Such findings about global and local processing will have wider implications in the real world. For example, these findings may assist those in the medical and scientific fields to understand how we process he information that we receive from our surrounding environment and if an accident were to affect our brain, how these processes would be altered or damaged.