Global Superiority Effect
- Pages: 9
- Word count: 2228
- Category: Corporation
A limited time offer! Get a custom sample essay written according to your requirements urgent 3h delivery guaranteedOrder Now
This study examined whether participant’s response times to global target were faster than local targets. Participants had to identify global and local shapes and letters as quick as possible and the response times which were recorded to the computer data. The study was a replication of Navon’s (1977), (cited in Ness Smith and Thirkettle, 2014) experiment but differed, as it was a focused attention task whereas this experiment was a divided attention task. This was the same as Yovel (2001), (cited in Ness, Smith and Thirkettle, 2014) mixed attention task study which tested the global to local accounts. The current experiment used a 2×2 repeated measures design, and all of the participants performed on all four conditions. Results showed that overall it took longer for participants to identify local stimuli compared to global stimuli.
A visual scene or any object in our environment contains both global features (whole) and local features (detailed). The question is do our eyes fixate on the global stimulus which is an overall image or do we perceive feature by feature in more detail? Navon (1977) cited in Ness et al. (2014), claimed that perceptual processes are temporarily organised, so that they proceed from global structuring towards more and finer grained analysis. His experiments aimed to show that perceptual systems process every scene starting with the global feature, leading to the local features. In Navon’s (1977) ‘focused attention tasks’ (cited in Ness et al. 2014), participants were directed to look at either the large global letter or the small local letters. He used large letters for his visual stimuli (the global level) which were made out of smaller letters (the local level). Participants had to identify either the larger characters or the smaller ones. The focus had to be globally or locally to stimuli that were consistent, neutral or conflicting on the global and local levels. The results revealed that the reaction time for global identification was much faster than local.
Even though Navon’s (1977), cited in Ness et al. (2014), theoretical predictions were correct, the support for his global to local account weakened. Yovel (2001) supported Navon’s (1977) claim that global letters were perceived prior to local letters, making the response times of participants faster (cited in Ness et al., 2014). However, he also suggested that if the component size for the local letter was larger, then there would be no difference between the reaction times to local and global target letters in the large component conditions. Yovel (2001) carried out a variation study of Navon’s (1977) original experiment (cited in Ness et al., 2014). Yovel’s (2001) was a ‘divided attention task’ which involved identifying the target letter at either global or local level, (the same in Navon’s study). However, the difference was that participants were asked to search for a target letter not knowing whether it would be a small local letter or large global letter. They had to search for targets on both levels; therefore, their attention was divided between both levels.
Only half the trials contained a target letter at either global or local level. The other non-target trials contained non-target letters at either local or global level. Yovel (2001) cited in Ness et al. (2014), hypothesised that there would be an interaction between target level and component size. He predicted no difference to reaction times on both global and local target letters in the large components condition, as the local letters were large, therefore equally as salient (cited in Ness et al., 2014). The results from Yovel’s study showed support for his prediction. When participants viewed the large components stimuli (local letters larger), the reaction times were faster. This showed an interaction between the two IV’s of target level and component size. Yovel (2001) cited in Ness et al. (2014), suggested that the findings highlighted the importance of stimulus characteristics for detecting perceptual differences. Also, the findings were consistent with the suggestion that ‘the equal salience of the local and global targets increase sensitivity of stimuli to the effects of how we perceive’ Yovel (2001) cited in Ness et al. (2014).
The current experiment is being run to investigate whether the global superiority effect is affected by the type of symbol, using a divided attention task. In previous studies Navon (1977), cited in Ness et al. (2014), used global letters made up of local letters to test how we process visual information. However, Yovel’s (2001), cited in Ness et al. (2014), divided attention task had two parts. Firstly, to see whether participants would react more quickly to global than local letters; and secondly, that there would be no difference to global and local target letters if the local letters were larger. In the current experiment the stimuli were global shapes and letters made up of local shapes and letters. They were tested to see the impact on processing and response times. The key features of the design was a 2×2 repeated measures design, and the participants performed on all four conditions which were two (IV’S) each with two levels. Three hypotheses were being tested. The first hypotheses was whether global is faster than local. The second hypothesis was if letters are faster than shapes. Finally, the third hypotheses was based on ‘Interaction’ which investigates whether the difference between global and local letters is more significant than the difference between global and local shapes. Method
The design used in the experiment was a two way repeated measure design (2×2). There were two Independent variables (IV) each with two levels. The first (IV) was ‘Target level’ with two levels ‘Local and Global’. The second (IV) was ‘Target Type’ and the two levels were ‘Letter and Shape’. There were four conditions in the experiment. The stimuli were larger global shapes and letters which were made up of smaller local shapes and letters. The participants had to search for a target letter which was an N aswell as a triangle shape. The dependent variable was the response time for the global letter/shape, and local letter/shape.
The capital letters presented to the participants included H,A,E,F and N, and the shapes that were presented were circle, square, diamond and triangle. The global shapes had 24 local components, either a single shape or letter repeated 24 times. All of the shapes and letters appeared at both global and local levels. The 100 stimuli were made using ‘Adobe Illustrator’. The participants had to press the button when the letter N or a triangle appeared on the screen at global or local level. The 4 stimuli containing a target at both levels were removed, and only 96 stimuli were shown to the participants in random order of 3 blocks of 32 in each. The stimuli were displayed for 120ms with an inter-stimulus interval of 2 seconds.
There were 101 participants of whom all were The Open University students. There was a mixture of male and female participants.
The stimulus presentation was carried out using E-Prime. The stimulus used in the experiment was Navon letters and shapes, they were created in Adobe Illustrator and presented using E-Prime.
“Response times to Global stimuli were, on average, 29ms faster than responses to Local stimuli. In addition, responses were 3 Ms faster for letter targets than for shape targets”.
F (IV DF, error DF) = F-ratio, p=sig
There was significant main effect of ‘Target level’ on participants reaction times (F1, 95) =7.38, p=.008) There was no significant main effect of ‘Target Type’ on participants reaction time (F, (1, 95) =0.1, p=0.82). Global letters were responded to more quickly than Global shapes or local targets. There was significant interaction between ‘Target Type and Target level’ (F, (1, 95) =6.9, p=.010)
The blue line represents Target Type (1=Letters) this slopes up dramatically, suggesting that reaction times were much slower for Target level 2 (Local Targets) than targets at the Global Level when the targets were letters.
The green line represents Target Type (2=Shapes), this suggests that reaction times were faster for Global Targets than local targets when targets were shapes.
Responses were faster for the letters than the shapes when the Target Type was global.
In the current experiment three hypothesis were being tested, Firstly ‘Global is faster than Local’. The results of the first hypothesis supported the prediction as there was a significant main effect on target levels on participant’s reaction times, showing the response for global stimuli to be 29ms faster than local stimuli. Similarly, Navon (1977), cited in Ness et al. (2014), also predicted in his study that there would be a faster response time for the processing of larger global letters than smaller local letters. Navon’s (1977) was a ‘focused attention task’, where participants were required to focus on either local or global level when they searched for a target letter. In the current experiment, participants performed in a ‘divided attention task’, where they were not aware if the target letter would appear at the local or global level. The results of the current experiment supported Navon’ s prediction, suggesting that the processing of global features would occur before the processing of local features.
Moreover, that the eyes initially focus on the bigger picture before we start noticing the surrounding and other fine details. However, Kinchla and Wolfe, (1979) argued that global processing would only occur prior to more detailed processing. The global structure can be ascertained by a single eye fixation which Navon’s (1977) study was unable to identify where in the visual processing system the global advantage occurs. The second hypothesis ‘letters are faster than shapes’ showed that there was no significance to the main effect of ‘Target Type’ on participant’s reaction time. Global letters were responded to more quickly than Global shapes or local targets. However, the mean response time for local shapes was faster than local letters. The reason for this could purely be that this may have occurred by chance. In previous research Navon (1977) or Yovel (2001) cited in Ness et al. (2014), did not use shapes as an independent variable.
But Navon did hypothesise that there would be an interaction between Target level and consistency (identifying a larger global letter ‘H’ made up of lots of small local letter S’s). In the current experiment, conflict or consistencies were not being measured. However, even though there was only a slight difference in response times which was 3 Ms faster for letter targets than for shape targets, participants attempted to recognise letters more quicker than symbols for which there may have been a delay. Shape could also be seen as conflict interfering with processing and slowing reaction times, as Navon (1977) had suggested. Also, Letters have semantic value which is used to understand human expression through language. Reading letters and words is a skill we are taught from a young age, therefore it may be easier for individuals to recognise and identify letters quicker. The third hypothesis was ‘Interaction’, “the difference between global and local letters is more significant than difference between global and local shapes”.
In the experiment Participants were faster at responding to global letters than shapes, when the targets were global. In contrast local letters were significantly slower than local shapes. The results showed significance to the hypothesis, suggesting that global was more significant than local. However, as before, local letters were slower than local shapes. This showed interaction occurred between local letter and local shape. Stimuli such as letters are more meaningful in comparison to shapes. In both stimuli, global advantage is observed although the interference effect only occurs with meaningful stimuli which increase reaction time for identification. This supports Navon’s (1977) theory of global precedence (cited in Ness et al., 2014). Global-local interference is a result of automatic processing of global targets. The global precedence effect has a sensory mechanism active in global advantage, whereas automatic semantic processes are active in the interference effect, indicating that it may occur by chance without attention or consciousness. It is possible that the variables tested may not reflect the psychological process that was being investigated.
As this was a divided attention task where either a letter or shape had to be identified, an effect observed in the results could be due to lack of concentration on task. There can be many reasons to this such as boredom or interruption which could essentially affect performance on reaction times. Moreover, in this experiment participants were allowed to practice for familiarity before the actual task itself. This could improve the participant’s results, as they would be mentally prepared in terms of expectations. However, this may be counterbalanced; for example, the participants could be split into two groups, mixture of females and males. The first group would experiment for ‘level’ first and then ‘type’ second. The second group would do ‘type’ first and ‘level’ second. The results would then be less affected by factors such as boredom and practice. In addition future research could look at what other factors may cause participants to focus more on either global or local stimuli. Research by Gasper and Clore (2002) found that participants in a lower mood are less likely to focus on the global figure. In addition, future research could also focus more on biological factors and what the causes may be, in relation to the brain and visual processing.