How we bisect lines and complex objects and how graspable objects result in larger biases Abstract The tendency of neurologically sound individuals to bisect lines and other stimuli slightly to the left has been studied greatly by various researchers

How we bisect lines and complex objects and how graspable objects result in larger biases

Abstract
The tendency of neurologically sound individuals to bisect lines and other stimuli slightly to the left has been studied greatly by various researchers (Fischer, 2000; Jewell & McCourt, 2000). Explanations for this include reading direction, attention and the activation-orientation hypothesis (Bultitude & Davies, 2006; S. Chokron, Bartolomeo, Perenin, Helft, & Imbert, 1998). However, few studies have investigated the effects of varying stimuli on bisection (Fischer, 2001). The present study investigates the difference between bisection error in lines and objects. We found a significant tendency to bisect graspable objects closer to the handle, however, found no pseudoneglect in objects or lines, or the differences between these stimuli. Further research should investigate graspable objects and bisection bias further with larger samples.

Introduction
Hemispatial neglect is the term used to explain the phenomena of patients with damage to the brain, typically following a stroke, who ignore one side of their visual field (Foulsham, Gray, Nasiopoulos, & Kingstone, 2013; Husain & Rorden, 2003). In most cases, damage to the right hemisphere of the brain, more specifically the right inferior parietal and temporoparietal lobe, results in patients being unaware of most stimulus in the left visual field, even though their vision is unaffected by the damage (Chokron, Colliot, & Bartolomeo, 2004; Husain & Rorden, 2003). Hemispatial neglect is frequently measured using line bisection tasks; whereby patients mark where they believe the centre of a line to be (Husain & Rorden, 2003; Jewell & McCourt, 2000). Patients with a neglect for the left side, following brain damage to the right hemisphere, will illustrate a significant rightward bias when bisecting lines as they do not perceive the left side of the line (Jewell & McCourt, 2000). Studies have noted that patients with hemispatial neglect tend to draw a clock with numbers only on the right side, however, found that by closing their eyes, patients drew the clock with no neglect (Chokron et al, 2004).
Interestingly, even neurologically healthy individuals show a bias, but instead of a rightward bias, they show a leftward bias (Bultitude & Davies, 2006; Fischer, 2000; Foulsham et al., 2013; Jewell & McCourt, 2000). This was first referred to as ‘pseudoneglect’ by Bowers & Heilman (as cited in Jewell & McCourt, 2000). Various studies have evidenced this effect, most finding a consistent leftward bias in healthy individuals, not only in lines but in images, words and simple shapes also (Bultitude & Davies, 2006; Fischer, 2000; Foulsham et al., 2013; Jewell & McCourt, 2000; Nielsen, Intriligator, & Barton, 1999). Bultitude & Davies (2006) even found pseudoneglect on a landmark test, where participants were shown lines that had already been bisected and had to indicate whether the mark was to the right or left of the centre. However, one study found no pseudoneglect when participants were able to move their gaze freely but when the stimulus was only perceived in one hemifield, the bias found was towards the fixation (Nielsen et al., 1999).
Explanations of pseudoneglect include the activation-orientation hypothesis (as cited in Bultitude & Davies, 2006), which explains that attention is biased in the visual field opposite to the most activated hemisphere. Furthermore, studies have found a link between reading direction and bias in line bisection (Chokron & Imbert, 1993). This study showed that Israeli participants (right to left readers), showed a rightward bias on a line bisection task, whereas French participants (left to right readers) showed a leftward bias (Chokron & Imbert, 1993). Additionally, a bias in line bisection can also be explained by the direction of scanning (Chokron, Bartolomeo, Perenin, Helft, & Imbert, 1998). In this study, participants were instructed to stop a moving marker on a line when it reached the centre; the line moved from either left to right or from right to left (Chokron et al, 1998). They found that there was a leftward bias for left to right scanning and a rightward bias for right to left scanning; the results were independent of reading direction (Chokron et al, 1998).
Although various studies, as discussed above, have investigated leftward bias in neurologically healthy participants, few have observed pseudoneglect in stimuli other than lines (Fischer, 2001; Foulsham et al., 2013). Although some have investigated pseudoneglect in words and relatively simple shapes, many studies have not observed pseudoneglect in complex objects (Fischer, 2000; Nielsen et al., 1999). Fischer (2001, p. 351) noted that to further research in pseudoneglect, experiments should, “specify stimulus attributes carefully.” Thus, our research, as well as observing pseudoneglect in lines, will also observe pseudoneglect in various objects, including objects that have a handle (graspable objects).
Our first research question is ‘do people show pseudoneglect in lines?’; we will investigate this by asking participants to bisect various lines throughout the experiment. Our second research question is ‘do people show pseudoneglect in objects?’; to test this question, we will include a variety of objects in our experiment and ask participants to bisect them horizontally. Our third research question is ‘is the amount of pseudoneglect different in different stimuli?’ whereby we will test by observing the differences of pseudoneglect between objects and lines. Finally, our fourth research question is ‘do people bisect graspable objects closer to the handle?’; for this research question, we will observe the error made by participants towards the handle of the graspable objects.
In relation to these research questions and the results of prior research, we hypothesise that for both lines and objects, participants will show a leftward bias in the bisection task. We also hypothesise that there will be a tendency for participants to bisect graspable objects closer to the handle as that is where we would pick up these objects up.

Methods
Participants
We obtained 13 participants through a voluntary basis for our experiment; 4 were female and 9 were male. All participants were right handed and had normal or corrected vision and ranged from 19 years old to 52 years old; the average age of the participants was 22.5 years old (SD=8.9).
Apparatus
Matlab (Mathworks) was used to run the experiment using Psychtoolbox. We used a variety of stimulus consisting of lines (a short line, 200 pixels horizontally, and a long line, 400 pixels horizontally), graspable objects (an axe, brush, tennis racket, saw, scissors, screwdriver, spanner and a spoon, all 400 pixels horizontally) and non-graspable objects (a bed, motorbike, bus, car, chest of drawers, fence, lorry and a sofa, all 400 pixels horizontally) (Rossion & Pourtois, 2004). We ensured that all objects selected were two-dimensional to ensure consistency throughout the stimulus.
Procedure
All participants were asked for verbal consent before taking part in the study. Participants were then directed to a computer where the experiment was ready to start; they were asked to note their age, sex and their dominant hand.
The experiment consisted of 192 trials in which various stimulus would appear on the screen in a random order. The stimulus was flipped horizontally and vertically in a random order and appeared on the screen in various places to ensure that participants did not use the screen itself to help bisect the stimulus. Furthermore, after each trial, the mouse returned to the centre of the screen so that the participant would not already have a bias to either the left or right of the screen. Participants had 2 seconds to bisect the stimulus on the screen where they thought the horizontal centre was. If they took longer than 2 seconds, they would receive an error message saying that they were too slow and they had to click the mouse to continue with the experiment. Once the experiment had ended, participants were thanked for their time and their results were recorded.

Results
To investigate the research question ‘do people show pseudoneglect in lines?’ and the hypothesis that participants would show a leftward bias, we conducted a one-sample t-test. No significant difference was found between the average error (in pixels) of lines (M=.15, SD=7.37), and the value ‘0’, t(12)=.071, p=.945, cohen’s d=.012, thus, the null hypothesis is not rejected.
To test the second research question ‘do people show pseudoneglect in objects?’ and the hypothesis that there would be a leftward bias, we used a one-sample t-test. This showed that there was no significant difference between the average error (in pixels) of objects (M=1.30, SD=4.78) and the value ‘0’, t(12)=.979, p=.347, cohen’s d=.272; we fail to reject the null hypothesis.
Furthermore, to test the third research question ‘is the amount of pseudoneglect different in different stimuli?’, we conducted a paired-samples t-test. The results of this t-test showed no significant difference between the average error (in pixels) of lines (M=.15, SD=7.37) and the average error (in pixels) of objects (M=1.30, SD=4.78), t(12)=.742, p=.472, cohen’s d=.206, so we fail to reject the null hypothesis.
Finally, to test the fourth research question ‘do people bisect graspable objects closer to the handle?’ and the hypothesis that graspable objects would be bisected more towards the handle, we first had to transform the data. We ‘reverse scored’ the data for ‘graspable objects’ when the image was flipped horizontally by multiplying the score by -1. This changes the data to now represent whether the error was closer to the handle of the object or not; all pictures of graspable objects had the handle to the right, thus, a positive number indicates a bias towards the handle of the object. We used a one-sample t-test to investigate this effect. The result showed that there is a significant difference between the average error (in pixels) towards the handle of graspable objects (M=2.93, SD=2.68) and the value ‘0’, t(12)=3.93, p=.002, cohen’s d=1.09, therefore we reject the null hypothesis.

Discussion
Our results show that although we failed to find pseudoneglect in lines and objects, and whether pseudoneglect is different between lines and objects, we found a significant result for a tendency to bisect graspable objects closer to the handle. Our results were in line with those found by Nielsen et al (1999), who found no pseudoneglect in normal bisection tasks. The only significant result found in our study was in relation to the research question ‘do people bisect graspable objects closer to the handle?’ in which we hypothesised that there would be a bias in bisection towards the handle. This could be because we almost alwars pick up graspable objects by the handle, so our attention is shifted to that side of the object, in preparation to pick it up. Previous studies have recognised the effect of attention in bisection, as found by Nicholls & Roberts (as cited in Bultitude & Davies, 2006).
Our results contradict many results in the field that have found a systematic leftward bias in neurologically healthy individuals (Fischer, 2000; Foulsham et al., 2013; Jewell & McCourt, 2000). Our differing results could be the result of various reasons. Firstly, our study consisted of a small sample of 13 participants; using more participants would allow us to see pseudoneglect, if any, much more clearly as it is difficult to draw conclusions from a small sample size. Additionally, due to such a small sample size, even with our significant result, it is too early to say with confidence that people bisect objects closer to the handle. To improve on future research, larger sample sizes should be used to investigate the effect of graspable objects on the bisection of those objects.
Furthermore, many of our participants were psychology students, who may have been informed about pseudoneglect; hence, may have tried to compensate for this effect by bisecting the line further right to their initial internal estimate. A further limitation of our study is that there were 192 continuous trials which could be boring for participants. This could affect the results as towards the end of the experiments, participants could become bored and their attention may be less focused on the task. To amend this, participants could be given a quick break half way through the trials in further studies to prevent fatigue effects.
Further research in the field could observe pseudoneglect in objects, specifically, different types of objects, as Fischer (2001) noted many experiments lacked. Building on our experiment, more research on the bisection of graspable and more complex objects should be conducted before any confident conclusions are made. This could provide insight into the ways in which we perceive objects and how this is useful in everyday life, for example, a bias towards the handles of objects may relate to the handle often being the safest place to pick up many graspable objects, such as a saw or a knife. The effect on the safety of graspable objects and where they are bisected in relation to their handle could additionally be investigated.
To conclude, our experiment shows some insight into potential biases in the bisection of graspable objects towards the handle. Research in this field could further investigate this effect by acquiring more participants in order to be more conclusive about the results of this study and potential explanations as to why we bisect graspable objects closer to the handle. Additionally, as we found no pseudoneglect in lines, as many studies have evidenced, future studies should aim to recruit participants who are unlikely to be informed on the topic of pseudoneglect. However, as our sample size was small, it may be too early to be certain about a tendency to bisect graspable objects closer to their handle. Nevertheless, our study provides some evidence towards the biases we have in bisecting complex objects.