View in EDS HTML Full Text PDF Full Text

The Effects of Spaces on Word Segmentation in Chinese Reading: Evidence from Eye Movements

Saved in:
Bibliographic Details
Title: The Effects of Spaces on Word Segmentation in Chinese Reading: Evidence from Eye Movements
Language: English
Authors: Liu, Pingping, Lu, Qin
Source: Journal of Research in Reading. May 2018 41(2):329-349.
Availability: Wiley-Blackwell. 350 Main Street, Malden, MA 02148. Tel: 800-835-6770; Tel: 781-388-8598; Fax: 781-388-8232; e-mail: cs-journals@wiley.com; Web site: http://www.wiley.com/WileyCDA
Peer Reviewed: Y
Page Count: 21
Publication Date: 2018
Document Type: Journal Articles
Reports - Research
Descriptors: Eye Movements, Chinese, Reading Processes, Orthographic Symbols, Language Processing, Phonemes, Reading Research
DOI: 10.1111/1467-9817.12106
ISSN: 0141-0423
Abstract: This paper studies the mechanisms behind the differential effects of inserting a space either before or after a two-character unit on information processing through the examination of eye movements in Chinese, a language where there is no word delimiters. A two-character unit in this study is either a word-preserving stimulus or a word-disrupting stimulus (i.e., nonword). The study aims for a better understanding of the cognitive mechanisms for lexical processing that may underlie observed facilitory or inhibitory effects of the spacing conditions in sentence context. Results show that inserting a space after a word facilitates lexical processing, but inserting a space before a word does not. Inserting a space before and after a nonword, however, does not show these effects. These results indicate that the effects of spaces before and after words are mainly influenced by word segmentation mechanisms rather than landing position effects or other factors.
Abstractor: As Provided
Entry Date: 2018
Accession Number: EJ1175754
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHWEQXTvPQYO98VqscxThWfAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDDGoJhqZXv7opyZDLQIBEICBm_HDmPQQWN4X8rK_M2N0fklZfW3ruuuAEnz9SpdZP4fTKGProT3BUNHbMOW6Z3-TJjGnTBDe-wVzRTJZv9zEqd3Fn_4k1ZSwSJFCD54te7ZwZTTqwU0uK4DxjaC0XZozWnSyHhAv2Yc5yIzhyBJaiNE9VL6M7Z0XoMvQbx88sFuaPC62AvP7u0SnPaKNp-6bMjArTaxKjVox0OEo
Text:
  Availability: 1
  Value: <anid>AN0129015805;d8c01may.18;2018Apr13.13:05;v2.2.500</anid> <title id="AN0129015805-1">The effects of spaces on word segmentation in Chinese reading: Evidence from eye movements </title> <p>This paper studies the mechanisms behind the differential effects of inserting a space either before or after a two‐character unit on information processing through the examination of eye movements in Chinese, a language where there is no word delimiters. A two‐character unit in this study is either a word‐preserving stimulus or a word‐disrupting stimulus (i.e., nonword). The study aims for a better understanding of the cognitive mechanisms for lexical processing that may underlie observed facilitory or inhibitory effects of the spacing conditions in sentence context. Results show that inserting a space after a word facilitates lexical processing, but inserting a space before a word does not. Inserting a space before and after a nonword, however, does not show these effects. These results indicate that the effects of spaces before and after words are mainly influenced by word segmentation mechanisms rather than landing position effects or other factors. <italic>What is already known about this topic</italic> Prior studies have found that differential effects of inserting a space either before or after a two‐character unit on information processing. However, the explanation for this finding is still unclear. <italic>What this paper adds</italic> This paper indicates that the differential effects of spaces before and after words are mainly caused by word segmentation mechanisms rather than landing position effects or other factors. <italic>Implications for theory, policy, or practice</italic> The present results are consistent with the assumptions of word segmentation and recognition model (Li, Rayner, & Cave, Cognitive Psychology, 2009). Furthermore, the implications for saccade target selection theories are also discussed.</p> <p>It is well known that spaces between words play an important role in most alphabetic writing systems because they aid word segmentation and eye guidance in reading. When spaces are deleted or masked, normal reading and eye movements are disrupted. Readers tend to make more and longer fixations, and land toward the word beginning when reading unspaced text compared with spaced text (McGowan, White, Jordan & Paterson, [<reflink idref="bib29" id="ref1">29</reflink>] ; Morris, Rayner & Pollatsek, [<reflink idref="bib31" id="ref2">31</reflink>] ; Perea & Acha, [<reflink idref="bib34" id="ref3">34</reflink>] ; Pollatsek & Rayner, [<reflink idref="bib36" id="ref4">36</reflink>] ; Rayner, Fischer & Pollatsek, [<reflink idref="bib38" id="ref5">38</reflink>] ; Rayner, Slattery & Belanger, [<reflink idref="bib39" id="ref6">39</reflink>] ; Rayner, Yang, Schuett & Slattery, [<reflink idref="bib40" id="ref7">40</reflink>] ; Winskel, Radach & Luksanneeyanawin, [<reflink idref="bib44" id="ref8">44</reflink>] ). However, there are no visual cues (e.g., spaces or other forms of demarcation) to mark boundaries between printed words in Chinese. If spaces are inserted in Chinese reading, how do they influence word segmentation and eye movements?</p> <p>Previous studies have found that inserting spaces between Chinese words does not typically speed up or disturb global sentence reading time[<reflink idref="bib6" id="ref9">6</reflink>] (Bai, Yan, Liversedge, Zang & Rayner, [<reflink idref="bib1" id="ref10">1</reflink>] ; Blythe et al., [<reflink idref="bib2" id="ref11">2</reflink>] ; Cui, Drieghe, Bai, Yan & Liversedge, [<reflink idref="bib7" id="ref12">7</reflink>] ; Inhoff, Liu, Wang & Fu, [<reflink idref="bib13" id="ref13">13</reflink>] ; I. M. Liu, Yeh, Wang & Chang, [<reflink idref="bib26" id="ref14">26</reflink>] ; Zang, Liang, Bai, Yan & Liversedge, [<reflink idref="bib48" id="ref15">48</reflink>] ). One reason is that inter‐word spaces have both inhibitory and facilitory effects on the processing of words in Chinese (Bai et al., [<reflink idref="bib1" id="ref16">1</reflink>] ; Blythe et al., [<reflink idref="bib2" id="ref17">2</reflink>] ; P. Liu & Li, [<reflink idref="bib22" id="ref18">22</reflink>] ; Zang et al., [<reflink idref="bib48" id="ref19">48</reflink>] ). Indeed, the visually unfamiliar format of word spaced text and the wider spatial distribution of spaced text may be disruptive to reading (Bai et al., [<reflink idref="bib1" id="ref20">1</reflink>] ; P. Liu & Li, [<reflink idref="bib22" id="ref21">22</reflink>] ). However, when available, inter‐word spaces delineate the spatial extent of Chinese words that are used to direct the eyes, and they segment character sequences into meaning‐conveying units (Inhoff & Wu, [<reflink idref="bib12" id="ref22">12</reflink>] ). In addition, these spaces between words may further aid the processing of words by reducing the effects of visual crowding (i.e., a reduced ability to identify objects in clutter) on eye movements (Bouma, [<reflink idref="bib3" id="ref23">3</reflink>] , [<reflink idref="bib4" id="ref24">4</reflink>] ; Brysbaert & Nazir, [<reflink idref="bib6" id="ref25">6</reflink>] ; McGowan, White & Paterson, [<reflink idref="bib30" id="ref26">30</reflink>] ; McGowan et al., [<reflink idref="bib29" id="ref27">29</reflink>] ; Perea & Acha, [<reflink idref="bib34" id="ref28">34</reflink>] ; Perea, Moret‐Tatay & Gomez, [<reflink idref="bib35" id="ref29">35</reflink>] ; Rayner et al., [<reflink idref="bib38" id="ref30">38</reflink>] ; Rayner et al., [<reflink idref="bib40" id="ref31">40</reflink>] ).</p> <p>Specifically, P. Liu and Li ([<reflink idref="bib22" id="ref32">22</reflink>] ) found that inserting a space after a word facilitates the processing of this word. However, they found that inserting a space before a word does not show any benefit, but sometimes even hinders the processing of this word. They suggested that the different effects of inserting a space before and after a word may be caused by word segmentation mechanisms based on the study by Li, Rayner and Cave's ([<reflink idref="bib20" id="ref33">20</reflink>] ) word segmentation and recognition model. In this model, Chinese word recognition is considered a sequential process, and only one word wins the competition although multiple character sequences are parallel activated initially in any given round. When a word is recognised, it is also segmented from the sentence, and the remaining characters will participate in the next round of competition. This model assumes that word recognition is simultaneous with the detection of word boundaries. Even though word recognition for a Chinese reader seems to be a simple task, the real challenge is to identify subsequences of characters that are both semantically and syntactically meaningful as word units (i.e., detecting where a word ends). Because most Chinese words are compounded, many words could begin with similar characters, and the reader needs to detect where the word or compound word ends. For example, when given the Chinese character sequence ‘研究生命’, it can either be segmented into 研究(explore)/生命(life) or 研究生(postgraduate)/命(destiny). The character ‘生’ can be the beginning of the word ‘生命’ to mean live or the end of this word ‘研究生’ to mean student. Under this context, the first segmentation is more reasonable. This so called overlapping ambiguity is very common in Chinese (Inhoff & Wu, [<reflink idref="bib12" id="ref34">12</reflink>] ; Li, Rayner & Cave, [<reflink idref="bib20" id="ref35">20</reflink>] ). Inserting a space after a word n marks the right boundary of word n (i.e., it tells the readers where the word ends); thus, this would facilitate the segmentation and recognition of word n from the rest of the sentence. As shown in Figure , when word n (呼吁, appealing) is segmented and recognised in normal unspaced text, both its left and right boundaries are known. Because the right boundary of word n is the left boundary of word n + 1, the left boundary (i.e., the beginning) of word n + 1 (社会, public) is also determined when word n is recognised. Thus, inserting a space before word n + 1 (i.e., a space after word n) cannot provide additional boundary information of word n + 1.</p> <p>Hypothetically speaking, in addition to word segmentation mechanisms, there may be a number of possible explanations to explain the different effects of inserting a space before and after a word, including differences in landing positions, the efficiency of parafoveal processing, visual crowding, the interruption of reading habit and so on. The second possibility for these different effects might be due to landing positions. The study by P. Liu and Li ([<reflink idref="bib22" id="ref36">22</reflink>] ) reported that readers initially tended to fixate on the first character (e.g., 呼) in the space after word condition (e.g., 呼吁) and on the second character (e.g., 吁) in the space before word condition (e.g.,呼吁). Previous studies of optimal viewing position (OVP) have also indicated that Chinese word recognition is most efficient when the eyes fixate more on the first character (i.e., OVP) than on the second character of a two‐character for both words and nonwords (P. Liu & Li, [<reflink idref="bib21" id="ref37">21</reflink>] ; P. Liu, Liu, Han & Paterson, [<reflink idref="bib23" id="ref38">23</reflink>] ). Therefore, the shorter reading time in the space after compared with the space before condition might be explained by initial fixations being more likely to be located at the OVP in the space after word condition.</p> <p>The third possible reason may be related to parafoveal processing. The different effects of inserting a space before and after a word may be a combination of parafoveal preview and parafoveal‐on‐foveal effects. Prior research has found that when readers have a valid preview of the word to the right of fixation, they spend less time fixating on that word compared with no provision of valid preview (e.g., Kennison & Clifton, [<reflink idref="bib15" id="ref39">15</reflink>] ; Rayner, [<reflink idref="bib37" id="ref40">37</reflink>] ; J. Yang, Li, Wang, Slattery & Rayner, [<reflink idref="bib47" id="ref41">47</reflink>] ). Some studies also found that the properties of a word in parafoveal vision have an immediate effect on foveal inspection time (e.g., low frequency of a word in parafoveal vision would increase foveal inspection time), and it is regarded as parafoveal‐on‐foveal effects (e.g., Cui et al., [<reflink idref="bib8" id="ref42">8</reflink>] ; Drieghe, [<reflink idref="bib9" id="ref43">9</reflink>] ; Kennedy & Pynte, [<reflink idref="bib14" id="ref44">14</reflink>] ). We found that inserting a space before word n + 1 cannot facilitate the processing of word n + 1, and the reason may be that this space extends the distance between word n + 1 and word n and thus would push word n + 1 further into the parafoveal region of reduced visual acuity. This may result in less parafoveal processing for word n + 1 compared with that in normal unspaced text. Consequently, it might reduce the preview benefit and thus inhibit the processing of word n + 1. In contrast, inserting a space after word n + 1 would not reduce the efficiency of parafoveal processing for this word compared with that of normal unspaced text, because it would have similar preview benefit effects and less parafoveal‐on‐foveal effects. Together, these parafoveal processing may explain why reading times were faster in the space after than the space before word condition.</p> <p>Fourthly, using space may also reduce visual crowding relative to normal unspaced text to facilitate word recognition (Bricolo, Salvi, Martelli, Arduino & Daini, [<reflink idref="bib5" id="ref45">5</reflink>] ; Brysbaert & Nazir, [<reflink idref="bib6" id="ref46">6</reflink>] ; McGowan et al., [<reflink idref="bib29" id="ref47">29</reflink>] ; Perea & Acha, [<reflink idref="bib34" id="ref48">34</reflink>] ; Perea et al., [<reflink idref="bib35" id="ref49">35</reflink>] ; Rayner et al., [<reflink idref="bib38" id="ref50">38</reflink>] ; Rayner et al., [<reflink idref="bib40" id="ref51">40</reflink>] ). The issue is whether the amount of reduction in visual crowding is identical in these two spacing conditions (i.e., before and after). Finally, inserting spaces in Chinese reading may actually disturb normal reading habit, although the level or amounts of disruption in different positions relative to a word may be different. Thus, the different effects of inserting a space in different positions to a word might be caused by several different factors rather than the demarcation of word boundaries only. This motivates us to study the major factors that cause the different effects of space use in Chinese text. We also examine saccade target selection in Chinese reading through spacing conditions because it is still unclear which factors determine where to move the eyes during Chinese reading (Li, Bicknell, Liu, Wei & Rayner, [<reflink idref="bib17" id="ref52">17</reflink>] ; Li, Liu & Rayner, [<reflink idref="bib18" id="ref53">18</reflink>] , [<reflink idref="bib19" id="ref54">19</reflink>] ; P. Liu, Li, Han & Li, [<reflink idref="bib25" id="ref55">25</reflink>] ; Y. Liu, Reichle & Li, [<reflink idref="bib24" id="ref56">24</reflink>] ; Ma, Li & Pollatsek, [<reflink idref="bib27" id="ref57">27</reflink>] ; Pan, Yan, Laubrock, Shu & Kliegl, [<reflink idref="bib33" id="ref58">33</reflink>] ; Shu, Zhou, Yan & Kliegl, [<reflink idref="bib41" id="ref59">41</reflink>] ; Yan, Kliegl, Richter, Nuthmann & Shu, [<reflink idref="bib45" id="ref60">45</reflink>] ; Zang et al., [<reflink idref="bib48" id="ref61">48</reflink>] ).</p> <p>In this study, we explore the effects of inserting spaces in different positions for two‐character units in Chinese word recognition using eye‐tracking as the mechanism. P. Liu and Li's ([<reflink idref="bib22" id="ref62">22</reflink>] ) research only studied space insertion between words. In the present study, the location of a space can be inserted either between words, referred to as word conditions, or within words, referred to as nonword (i.e., word‐disrupting) conditions, respectively. In the space before word (or nonword) condition, a space was inserted before the target stimulus; in the space after word (or nonword) condition, a space was inserted after the target stimulus. In the spaces around word (or nonword) condition, there was a space before the target stimulus and another space after it. Table  shows the five explored factors, and the columns show the hypothesised effects of space to these factors. First, the key difference between these words and nonwords is related to word segmentation. For nonword stimuli, the inhibitory effect of spaces on word segmentation should be similar between the space before and the space after the nonword condition because a word was equally disrupted by a space regardless of the location of this space. Secondly, according to the OVP effects, there should be a shorter time in the space after compared with the space before condition for both words and nonwords because initial fixations are more likely to be located at the OVP in the space after word condition. Thirdly, the explanation of parafoveal processing should have a facilitatory effect in the space after compared with the space before condition for both words and nonwords because of the preview benefit and less parafoveal‐on‐foveal effects. Fourthly, compared with the normal unspaced text, these spaces between words or nonwords might aid information processing because of the amount of reduction in visual crowding. Finally, the inhibitory effects of spaces on reading habit may be similar between the word and nonword conditions. Accordingly, if different effects of inserting a space before and after a word might be caused by factors other than word segmentation, there would be differences in the effects of inserting a space before or after this nonword stimulus. Otherwise, these different effects of inserting spaces before and after words could be mainly caused by word segmentation mechanisms, which could be consistent with the predictions of the word segmentation and recognition model (Li et al., [<reflink idref="bib20" id="ref63">20</reflink>] ).</p> <p>List of possible facilitatory and inhibitory effects of spaces on information processing relative to the normal unspaced text in Chinese</p> <p> <ephtml> <table border="1" cellpadding="7"><tr><th /><th>Word = word‐preserving unit</th><th>Nonword = word‐disrupting unit</th></tr><tr><th>Example</th><th>呼吁</th><th>呼吁</th><th>呼吁</th><th>吁社</th><th>吁社</th><th>吁社</th></tr><tr><td>Factors</td><td> Space before</td><td> Space after</td><td> Space around</td><td> Space before</td><td> Space after</td><td> Space around</td></tr><tr><td>Word segmentation</td><td>×</td><td>√</td><td>√</td><td>×</td><td>×</td><td>×</td></tr><tr><td>Landing position</td><td>×</td><td>√</td><td>–</td><td>×</td><td>√</td><td>–</td></tr><tr><td>Parafoveal processing</td><td>×</td><td>√</td><td>√ or –</td><td>×</td><td>√</td><td>√ or –</td></tr><tr><td>Visual crowding</td><td>√</td><td>√</td><td>√</td><td>√</td><td>√</td><td>√</td></tr><tr><td>Reading habit</td><td>×</td><td>×</td><td>×</td><td>×</td><td>×</td><td>×</td></tr></table> </ephtml> </p> <p>1 Note: The symbol “√” indicates a facilitatory effect, the symbol “×” indicates an inhibitory effect, and the symbol “ –” indicates neither a facilitatory nor an inhibitory effect. The space character is indicated by the shaded box which is blank space in the experiment.</p> <hd id="AN0129015805-2">Method</hd> <hd id="AN0129015805-3">Participants</hd> <p>Forty native Chinese speakers (20 women; average age = 22.4 years; range = 18.3–25.8 years) from several universities near the Institute of Psychology in Beijing were paid to participate in the experiment. All participants were proficient in Chinese, had normal or corrected‐to‐normal vision and were unaware of the purpose of the experiment.</p> <hd id="AN0129015805-4">Apparatus</hd> <p>Eye movements were recorded using an Eyelink 1,000 tower‐mounted system (SR Research Ltd, Osgoode, Canada). Viewing was binocular, but only the right eye was monitored. The sentences were displayed on a 21‐inch CRT monitor (resolution: 1024 × 768 pixels; refresh rate: 150 Hz) connected to a Dell PC. Participants were seated approximately 58 cm away from the computer monitor, and each character subtended a visual angle of approximately 1.2°. Each sentence was presented on a single line with the Song 24‐point font. The size of each space was 16 × 32 pixels (width by height) in the spacing condition.</p> <hd id="AN0129015805-5">Materials and design</hd> <p>Eighty experimental sentences were developed for this experiment, and they were 20–28 characters in length (M = 23.0, SD = 1.7). Acceptability ratings for each sentence were obtained from 13 native speakers of Chinese (None of whom participated in the main experiment) on a scale of 1 (totally unacceptable) to 7 (perfectly acceptable). The average reported naturalness was 6.40 (SD = .33).</p> <p>As stated earlier, there were eight presentation conditions for each experimental sentence frame (Table ). Besides the normal unspaced condition, there were three spacing conditions for both word (e.g., 呼吁, appealing) and nonword stimuli (e.g., 吁社, which was a sequence of two characters including the second character of one word and the first character of the next): a space before the stimulus, a space after the stimulus and spaces around (both before and after) the stimulus. Additionally, there was a space after four characters condition, which was included as a filler condition. The four‐character region (e.g., 呼吁社会, appealing to the public) consists of two two‐character words, which were listed as words in the Chinese Lexicon ([<reflink idref="bib16" id="ref64">16</reflink>] ). Out of 80 words in the first position (e.g., 呼吁), which were chosen randomly, 34 are common nouns, 22 are verbs, 14 are adverbs, 8 are adjectives and 2 are pronouns. As shown in Table , the four characters were matched for character complexity, character frequency and orthographic neighbourhood size (Zhou, [<reflink idref="bib49" id="ref65">49</reflink>] ) (all p values > .50). In order to avoid the confounding effects of sentence beginning or end (e.g., wrap‐up effect) on target word processing, these four characters were always in the middle of the experimental sentences and were not within the first five or last five characters of a sentence. A counterbalanced design was used to enable each participant to view an equal number of sentences in each condition but read only one sentence for each item. We constructed eight lists of materials such that 10 items appeared in each condition in each list and each item appeared in a different condition in each list. Additionally, eighty sentences with spaces were included as fillers to prevent participants from predicting the positions of spaces within the experimental materials. Thus, in the main experiment, each participant reads 16 practice sentences, 80 experimental items and 80 filler sentences.</p> <p>Example sentences with inserted spaces.</p> <p> <ephtml> <table border="1" cellpadding="2"><tr><th>Condition</th><th>Sentence</th></tr><tr><td>Normal unspaced condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space before the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space after the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Spaces around the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space after four characters (filler)</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space before the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space after the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Spaces around the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr></table> </ephtml> </p> <p>2 Note: The English translation of the sentence is ‘This charity organisation has been appealing to the public to pay attention to children with mental retardation for many years’. The four‐character interest of region ‘呼吁社会’ (i.e., appealing to the public) has been highlighted by bold font (but not during the actual experiment), and the space character is indicated by the shaded box which is blank space in the experiment.</p> <p>Properties of the four characters used in the study.</p> <p> <ephtml> <table border="1" cellpadding="7"><tr><th /><th>First character</th><th>Second character</th><th>Third character</th><th>Fourth character</th><th /><th /></tr><tr><th>Example</th><th>呼</th><th>吁</th><th>社</th><th>会</th><th>F</th><th>p</th></tr><tr><td>Character frequency</td><td>1,496 (164)</td><td>1,353 (156)</td><td>1,307 (157)</td><td>1,469 (180)</td><td>.36</td><td>.78</td></tr><tr><td>Character complexity</td><td>8.0 (.32)</td><td>8.5 (.27)</td><td>8.0 (.27)</td><td>8.1 (.29)</td><td>.68</td><td>.57</td></tr><tr><td>Orthographic neighbourhood size</td><td>8.85 (.68)</td><td>8.80 (.68)</td><td>8.35 (.61)</td><td>8.65 (.65)</td><td>.12</td><td>.73</td></tr></table> </ephtml> </p> <p>3 Note: Numbers in parentheses are standard errors. Character frequency is in occurrences per million. The number of individual strokes in a character is treated as the index of visual complexity. The degrees of freedom were 3 and 237.</p> <hd id="AN0129015805-6">Procedure</hd> <p>Participants were tested individually. After participants arrived at the lab, they were given instructions for the experiment and a description of the apparatus. At the start of the experiment, participants performed a calibration procedure by looking at a sequence of three fixation points randomly displayed horizontally across the middle of the computer screen. Following the commonly practiced calibration principles for Chinese character, the maximum gaze position error was smaller than 0.5° of the visual angle. In this experiment, the actual gaze position error never exceeded 0.3°, and the average gaze position error was smaller than 0.2° of the visual angle. At the beginning of each trial, a white square (approximately 1° × 1°) appeared on the left side of the computer screen, which indicated the position of the first character in the sentence. Once the participant successfully fixated on the white square, a sentence was presented. Participants were instructed to read silently for comprehension, and they were told that they would periodically be asked to answer yes–no questions that appeared following 37% of the sentences and were related to the meaning of the sentence. These yes or no questions required the participants to have understood the meaning of the sentence and respond via a button press. The eye tracker was checked and recalibrated if necessary prior to the presentation of each sentence. Participants were given short breaks between trials to prevent fatigue, and the experiment lasted approximately 40 minutes.</p> <hd id="AN0129015805-7">Results and discussion</hd> <p>The average comprehension accuracy was 95%, and there were no significant differences across these experimental conditions. Trials in which a blink was made before or after a fixation on the four‐character region were discarded prior to analyses (6.2% of trials). Fixations were also removed if they were shorter than 80 ms or longer than 1000 ms (0.6% of fixations).</p> <p>A series of standard eye movement measures for the two‐character region of interest (ROI, Table ) were reported: (a) first fixation duration (the duration of the first fixation on the ROI during the first‐pass reading), (b) gaze duration (the sum of all first‐pass fixations on the ROI before moving to another word), (c) total time (the sum of all fixations on the ROI, including fixations that follow a regression), (d) fixation probability (the probability of fixating on the ROI during the first‐pass reading) and (e) initial landing positions (the position of the first fixation on a word). In order to show the effects of spaces on word segmentation and eye movements clearly, a series of 2 (stimulus type: word versus nonword) × 4 (spacing condition: normal, before, after and around) repeated measures of ANOVAs (i.e., analyses of variance) were undertaken for both participants' (F<subs>1</subs>) and items' (F<subs>2</subs>) means. In order to reduce false positives in this experiment (von der Malsburg & Angele, [<reflink idref="bib43" id="ref66">43</reflink>] ), Bonferroni correction was applied to these multiple comparisons between the different spacing conditions if the main effects were significant. The factor of parafoveal processing can be tested by the comparison between the space before the nonword condition and the space after the nonword condition, because the amount of reduction in visual crowding and reading habits might be similar between the two conditions. The factor of visual crowding effect can be tested by the comparison between the space before the nonword condition and the spaces around the nonword condition, because the amount of reduction in parafoveal processing and reading habits might be similar between the two conditions.</p> <p>Example of regions of interest for two characters</p> <p> <ephtml> <table border="1" cellpadding="2"><tr><th>Condition</th><th>Two characters</th></tr><tr><td>Normal unspaced condition (word)</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space before the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space after the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Spaces around the word condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Normal unspaced condition (nonword)</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space before the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Space after the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr><tr><td>Spaces around the nonword condition</td><td>这家慈善机构多年来一直呼吁社会关心智障儿童。</td></tr></table> </ephtml> </p> <p>4 Notes. The two‐character interest of region has been highlighted by bold font (but not during the actual experiment), and the space character is indicated by the shaded box which is blank space in the experiment.</p> <hd id="AN0129015805-8">First fixation duration</hd> <p>Figure  shows the results of the first fixation duration. The significant main effects of spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref67">3</reflink>, 117) = 3.69, MSE = 1,651, p = .01, η<subs>p</subs><sups>2</sups> = .09, F<subs>2</subs>(<reflink idref="bib3" id="ref68">3</reflink>, 237) = 6.12, MSE = 2,848, p = .001, η<subs>p</subs><sups>2</sups> = .07, and stimulus type, F<subs>1</subs>(<reflink idref="bib1" id="ref69">1</reflink>, 38) = 4.63, MSE = 910, p = .04, η<subs>p</subs><sups>2</sups> = .11, F<subs>2</subs>(<reflink idref="bib1" id="ref70">1</reflink>, 79) = 11.57, MSE = 1,590, p = .001, η<subs>p</subs><sups>2</sups> = .13, were qualified by a significant interaction, F<subs>1</subs>(<reflink idref="bib3" id="ref71">3</reflink>, 117) = 2.86, MSE = 1,055, p = .04, η<subs>p</subs><sups>2</sups> = .07, F<subs>2</subs>(<reflink idref="bib3" id="ref72">3</reflink>, 237) = 3.71, MSE = 2,211, p = .01, η<subs>p</subs><sups>2</sups> = .04. Further analyses showed that spacing type did have strong effects on first fixation durations in the word condition, F<subs>1</subs>(<reflink idref="bib3" id="ref73">3</reflink>, 117) = 5.81, MSE = 1,429, p = .001, η<subs>p</subs><sups>2</sups> = .13, F<subs>2</subs>(<reflink idref="bib3" id="ref74">3</reflink>, 237) = 9.60, MSE = 2,310, p < .001, η<subs>p</subs><sups>2</sups> = .11, but did not in the nonword condition, F<subs>1</subs>(<reflink idref="bib3" id="ref75">3</reflink>, 117) = .63, p = .60, F<subs>2</subs>(<reflink idref="bib3" id="ref76">3</reflink>, 237) = 1.26, p = .29. Follow‐up contrasts indicated that first fixation durations were significantly longer in the space before the word condition (268 ms) than in the space after the word condition (239 ms) and the spaces around the word condition (237 ms; all p values < .01). There was a significant difference between the normal unspaced condition (253 ms) and the space before the word condition (268 ms) by items, F<subs>2</subs>(<reflink idref="bib1" id="ref77">1</reflink>, 79) = 8.25, MSE = 4,570, p = .005, η<subs>p</subs><sups>2</sups> = .095, but not by participants, F<subs>1</subs>(<reflink idref="bib1" id="ref78">1</reflink>, 39) = 2.11, MSE = 4,296, p = .16, η<subs>p</subs><sups>2</sups> = .051. These results indicate that inserting a space after a stimulus facilitated information processing in the word condition, but not in the nonword condition. Furthermore, the nonsignificant differences between the space before and the space after the nonword conditions indicate that parafoveal processing may not play a key role in the different effects of spacing locations on word recognition.</p> <hd id="AN0129015805-9">Gaze duration</hd> <p>Figure  shows the result of the gaze duration. There was a significant effect of spacing condition on gaze duration, F<subs>1</subs>(<reflink idref="bib3" id="ref79">3</reflink>, 117) = 3.75, MSE = 3,398, p = .01, η<subs>p</subs><sups>2</sups> = .09, F<subs>2</subs>(<reflink idref="bib3" id="ref80">3</reflink>, 237) = 5.71, MSE = 5,607, p = .001, η<subs>p</subs><sups>2</sups> = .07. There was also a significant effect of stimulus type, F<subs>1</subs>(<reflink idref="bib1" id="ref81">1</reflink>, 39) = 13.63, MSE = 2,434, p = .001, η<subs>p</subs><sups>2</sups> = .26; F<subs>2</subs>(<reflink idref="bib1" id="ref82">1</reflink>, 79) = 14.83, MSE = 5,885, p < .001, η<subs>p</subs><sups>2</sups> = .16, as participants fixated longer on stimuli of nonword (293 ms) than word (273 ms). The interaction between stimulus type and spacing condition was not significant (p values > .10). Further analysis showed that spacing condition influenced gaze duration significantly in the word condition, F<subs>1</subs>(<reflink idref="bib3" id="ref83">3</reflink>, 117) = 4.78, MSE = 2,609, p = .004, η<subs>p</subs><sups>2</sups> = .11, F<subs>2</subs>(<reflink idref="bib3" id="ref84">3</reflink>, 237) = 7.68, MSE = 4,586, p < .001, η<subs>p</subs><sups>2</sups> = .09, but not in the nonword condition, F<subs>1</subs>(<reflink idref="bib3" id="ref85">3</reflink>, 117) = .76, p = .52, F<subs>2</subs>(<reflink idref="bib3" id="ref86">3</reflink>, 237) = 1.09, p = .35. Follow‐up contrasts showed that gaze duration was significantly longer in the space before the word condition (291 ms) than in the space after the word condition (253 ms) and the spaces around the word condition (262 ms; all p values < .05). There was no significant difference between the normal (284 ms) unspaced condition and the space before (291 ms) the word condition, F<subs>1</subs>(<reflink idref="bib1" id="ref87">1</reflink>, 39) = .20, p = .66, F<subs>2</subs>(<reflink idref="bib1" id="ref88">1</reflink>, 79) = .76, p = .39. More specifically, the nonsignificant differences between the space before and the spaces around the nonword conditions indicated that visual crowding is not likely to play a key role in the different effects of spacing locations on word recognition. In summary, we found disparate results between word and nonword stimuli for both first fixation durations and gaze durations.</p> <p>In order to test whether the different effects of inserting a space before and after a word are influenced by landing positions or not, we examined the gaze duration when readers initially fixated at either the first or the second character of the word. Prior findings of the OVP effect indicated that Chinese word recognition is more efficient when the eyes fixate on the first character than on the second character of an isolated two‐character word (P. Liu & Li, [<reflink idref="bib21" id="ref89">21</reflink>] ; P. Liu et al., [<reflink idref="bib19" id="ref90">19</reflink>] ). If landing position plays a key role in the different effects of spacing locations on word recognition, there should be shorter gaze duration when the eyes fixated on the first character than on the second character in these spacing conditions. However, the following results did not support this assumption. Results showed that Chinese readers did not recognise a word more efficiently when they fixated on the first character than on the second character within a word in the normal unspaced condition, t<subs>1</subs>(<reflink idref="bib1" id="ref91">1</reflink>, 39) = 1.64, p = .11, t<subs>2</subs>(<reflink idref="bib1" id="ref92">1</reflink>, 67) = 1.96, p = .054, or in the spaces around the word condition, t<subs>1</subs>(<reflink idref="bib1" id="ref93">1</reflink>, 36) = .32, p = .75, t<subs>2</subs>(<reflink idref="bib1" id="ref94">1</reflink>, 65) = .94, p = .35. More specifically, in the space after the word condition, gaze duration was longer when the eyes fixated on the first character (271 ms) than on the second character (223 ms) within a word, t<subs>1</subs>(<reflink idref="bib1" id="ref95">1</reflink>, 39) = 3.56, p = .001, t<subs>2</subs>(<reflink idref="bib1" id="ref96">1</reflink>, 57) = 2.64, p = .011. In contrast, in the space before the word condition, gaze duration was shorter when the eyes fixated on the first character (271 ms) than on the second character position (314 ms) within a word, t<subs>1</subs>(<reflink idref="bib1" id="ref97">1</reflink>, 33) = 2.17, p = .038, t<subs>2</subs>(<reflink idref="bib1" id="ref98">1</reflink>, 60) = 2.03, p = .047. Actually, the results indicate that word recognition was more efficient in the space after the word condition than the space before the word condition. This group of results is not consistent with the assumption of landing position, which was not considered in the further analyses.</p> <p>Furthermore, multiple linear regression analyses were performed to assess how these factors (i.e., word segmentation, parafoveal processing, crowding and reading habits) influence the different effects of inserting a space before and after a word. In the analysis, the dependent variable was extracted by the average differences of gaze duration in the spacing condition and in the normal unspaced condition. Because it is difficult to obtain the exact value of these factors, we follow the same logical value assignment principle in prior studies (e.g., Bai et al., [<reflink idref="bib1" id="ref99">1</reflink>] ; Blythe et al., [<reflink idref="bib2" id="ref100">2</reflink>] ; Cui et al., [<reflink idref="bib7" id="ref101">7</reflink>] ; Cui et al., [<reflink idref="bib8" id="ref102">8</reflink>] ; McGowan et al., [<reflink idref="bib30" id="ref103">30</reflink>] ), we assigned the maximum facilitatory effect to have the value of positive one (<reflink idref="bib1" id="ref104">1</reflink>), the maximum inhibitory effect to have the value of negative one (−1). The effects of word segmentation were coded as positive one (<reflink idref="bib1" id="ref105">1</reflink>) in the space after and around the word condition and negative one (−1) in other spacing conditions. The effects of parafoveal processing in the space before, after and around the stimuli were coded as −1, 0.5 and 0.3, respectively. The effects of crowding in the space before, after and around the stimuli were coded as 0.5, 0.5 and 1, respectively. The effects of reading habits were coded as −1 in the spaces around nonword condition and were coded as −0.5 in other spacing conditions. The values of these factors were treated as independent variables to determine which one has stronger influence in the different effects of spacing. We used a stepwise method to examine the predictive contribution of these factors. All factors significant at the p < 0.05 level were entered into the final model, R<sups>2</sups> = 0.017; F(<reflink idref="bib1" id="ref106">1</reflink>, 1,<reflink idref="bib751" id="ref107">751</reflink>) = 30.87, p < .001. Results showed that the contribution of word segmentation was significant (B‐score = −19.65, SE = 3.54, standardised B‐values = −.13, t = −5.56, p < .001), but the contribution of parafoveal processing (p = .18), crowding (p = .75) or reading habits (p = .94) was not significant. In short, these different effects of inserting a space before and after a word are more likely to be influenced by word segmentation mechanisms.</p> <hd id="AN0129015805-10">Total time</hd> <p>Figure  shows the result of the total time. There was a significant main effect of stimulus type, F<subs>1</subs>(<reflink idref="bib1" id="ref108">1</reflink>, 39) = 13.99, MSE = 8,385, p = .001, η<subs>p</subs><sups>2</sups> = .26; F<subs>2</subs>(<reflink idref="bib1" id="ref109">1</reflink>, 79) = 16.30, MSE = 11,446, p < .001, η<subs>p</subs><sups>2</sups> = .17, as participants fixated longer on stimuli of nonword (378 ms) than word (340 ms). When participants fixated their eyes within that nonword stimulus, they might have to process two words simultaneously, which is a difficult task and lead to longer fixation durations. There was also a significant interaction between stimulus type and spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref110">3</reflink>, 117) = 3.22, MSE = 5,488, p = .03, η<subs>p</subs><sups>2</sups> = .08, F<subs>2</subs>(<reflink idref="bib3" id="ref111">3</reflink>, 237) = 6.60, MSE = 8,770, p < .001, η<subs>p</subs><sups>2</sups> = .08. Further analyses showed that spacing significantly influenced total time in the word condition, F<subs>1</subs>(<reflink idref="bib3" id="ref112">3</reflink>, 117) = 2.87, MSE = 5,376, p = .04, η<subs>p</subs><sups>2</sups> = .07, F<subs>2</subs>(<reflink idref="bib3" id="ref113">3</reflink>, 237) = 4.32, MSE = 9,835, p = .005, η<subs>p</subs><sups>2</sups> = .05, but only by item in the nonword condition, F<subs>1</subs>(<reflink idref="bib3" id="ref114">3</reflink>, 117) = 1.31, MSE = 9,150, p = .28, η<subs>p</subs><sups>2</sups> = .03, F<subs>2</subs>(<reflink idref="bib3" id="ref115">3</reflink>, 237) = 3.26, MSE = 13,378, p = .02, η<subs>p</subs><sups>2</sups> = .04. Follow‐up contrasts indicated that total time was significantly longer in the space before the word condition (358 ms) than in the space after the word condition (321 ms) and in the spaces around the word condition (324 ms; all p values < .05). Total time did not significantly differ, F<subs>1</subs>(<reflink idref="bib1" id="ref116">1</reflink>, 39) = .02, p = .89, F<subs>2</subs>(<reflink idref="bib1" id="ref117">1</reflink>, 79) = .63, p = .43, between the normal unspaced condition (355 ms) and the space before the word condition (358 ms). In contrast, for the nonword condition, total time was longer in the spaces around the nonword condition (403 ms) than in the normal (367 ms) unspaced condition, F<subs>1</subs>(<reflink idref="bib1" id="ref118">1</reflink>, 39) = 3.60, p = .07, F<subs>2</subs>(<reflink idref="bib1" id="ref119">1</reflink>, 79) = 6.66, p = .01, in the space before (378 ms) the nonword condition, F<subs>1</subs>(<reflink idref="bib1" id="ref120">1</reflink>, 39) = 1.14, p = .29, F<subs>2</subs>(<reflink idref="bib1" id="ref121">1</reflink>, 79) = 3.98, p = .05, and in the space after (365 ms) the nonword condition, F<subs>1</subs>(<reflink idref="bib1" id="ref122">1</reflink>, 39) = 2.98, p = .09, F<subs>2</subs>(<reflink idref="bib1" id="ref123">1</reflink>, 79) = 8.00, p = .006. Consistent with prior studies (Bai et al., [<reflink idref="bib1" id="ref124">1</reflink>] ; P. Liu & Li, [<reflink idref="bib22" id="ref125">22</reflink>] ; Zang et al., [<reflink idref="bib48" id="ref126">48</reflink>] ), inserting a space within a word interfered with information processing. These results also showed that inserting spaces before and after word stimuli affects information processing differently, but inserting spaces before and after nonword stimuli does not show these effects.</p> <hd id="AN0129015805-11">Fixation probability</hd> <p>There was a significant main effect of stimulus type, as participants fixated more frequently on stimuli of nonword (.82) than on word (.77), F<subs>1</subs>(<reflink idref="bib1" id="ref127">1</reflink>, 39) = 21.06, MSE = 134, p < .001, η<subs>p</subs><sups>2</sups> = .35, F<subs>2</subs>(<reflink idref="bib1" id="ref128">1</reflink>, 79) = 9.18, MSE = 545, p = .003, η<subs>p</subs><sups>2</sups> = .10. There was also a significant main effect of spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref129">3</reflink>, 117) = 4.30, MSE = 369, p = .006, η<subs>p</subs><sups>2</sups> = .10, F<subs>2</subs>(<reflink idref="bib3" id="ref130">3</reflink>, 237) = 9.36, MSE = 349, p < .001, η<subs>p</subs><sups>2</sups> = .11, as well as a significant interaction between stimulus type and spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref131">3</reflink>, 117) = 3.31, MSE = 186, p = .02, η<subs>p</subs><sups>2</sups> = .08; F<subs>2</subs>(<reflink idref="bib3" id="ref132">3</reflink>, 237) = 3.02, MSE = 352, p = .03, η<subs>p</subs><sups>2</sups> = .04. For the word conditions, follow‐up contrasts indicated that participants were more likely to make a fixation on these words in the normal unspaced condition (.84) than in the space before the word condition (.71) and the spaces around the word condition (.74; all p values < .05). For the nonword conditions, participants were less likely to fixate on these stimuli in the space before the nonword condition (.76) than in the normal unspaced condition (.83), in the space after the nonword condition (.86), or the spaces around the nonword condition (.85; all p values < .01).</p> <hd id="AN0129015805-12">Initial landing position</hd> <p>Figure  shows the results of the initial landing position. There was a significant main effect of spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref133">3</reflink>, 117) = 24.43, MSE = .04, p < .001, η<subs>p</subs><sups>2</sups> = .39, F<subs>2</subs>(<reflink idref="bib3" id="ref134">3</reflink>, 237) = 34.49, MSE = .06, p < .001, η<subs>p</subs><sups>2</sups> = .30. Follow‐up contrasts showed that initial landing positions in the space after the stimulus condition (.38) were significantly closer to the beginning of the target region than that in the other three conditions (all p values < .05). Initial landing positions in the normal unspaced condition (.44) were significantly closer to the beginning of the target region than that in the space before the stimulus (.61) and the spaces around the stimulus conditions (.55; all p values < .05). There was no interaction between stimulus type and spacing condition (all p values > .60), nor a significant main effect of stimulus type (all p values > .80), indicating that this effect held identically for both words and nonwords. Thus, these results show that inserting spaces into text produced a shift in landing positions away from the space regardless of the relative location of the space within or between words.</p> <hd id="AN0129015805-13">Initial landing position for single fixations</hd> <p>To further test how Chinese readers select saccade target during reading, initial landing positions for single fixations are further analyzed. Table  shows the proportions of single fixations and multiple fixations cases during first‐pass reading across all trials.[<reflink idref="bib7" id="ref135">7</reflink>] The results were similar to those reported earlier for initial landing positions regardless of the number of fixations made. As shown in Figure , there was no hint of a significant interaction between stimulus type and spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref136">3</reflink>, 117) = .31, p = .82, F<subs>2</subs>(<reflink idref="bib3" id="ref137">3</reflink>, 237) = .26, p = .86, nor a significant main effect of stimulus type, F<subs>1</subs>(<reflink idref="bib1" id="ref138">1</reflink>, 39) = .24, p = .63, F<subs>2</subs>(<reflink idref="bib1" id="ref139">1</reflink>, 79) = .19, p = .67.</p> <p>The proportions of single fixations and multiple fixations cases during first‐pass reading for all trials.</p> <p> <ephtml> <table border="1" cellpadding="4"><tr><th>Condition</th><th>Single fixations</th><th>Multiple fixations</th><th>Skip</th></tr><tr><td>Normal unspaced condition (word)</td><td>0.72</td><td>0.00</td><td>.28</td></tr><tr><td>Space before the word condition</td><td>0.65</td><td>0.07</td><td>.28</td></tr><tr><td>Space after the word condition</td><td>0.72</td><td>0.05</td><td>.23</td></tr><tr><td>Spaces around the word condition</td><td>0.67</td><td>0.07</td><td>.26</td></tr><tr><td>Normal unspaced condition (nonword)</td><td>0.71</td><td>0.12</td><td>.17</td></tr><tr><td>Space before the nonword condition</td><td>0.63</td><td>0.13</td><td>.24</td></tr><tr><td>Space after the nonword condition</td><td>0.73</td><td>0.13</td><td>.14</td></tr><tr><td>Spaces around the nonword condition</td><td>0.74</td><td>0.11</td><td>.15</td></tr><tr><td>Mean</td><td>0.70</td><td>0.10</td><td>.20</td></tr></table> </ephtml> </p> <p>There was, however, a significant main effect of spacing condition, F<subs>1</subs>(<reflink idref="bib3" id="ref140">3</reflink>, 117) = 17.30, MSE = .04, p < .001, η<subs>p</subs><sups>2</sups> = .31, F<subs>2</subs>(<reflink idref="bib3" id="ref141">3</reflink>, 237) = 22.22, MSE = .07, p < .001, η<subs>p</subs><sups>2</sups> = .22. Follow‐up contrasts showed that initial landing positions for single fixations in the space after the stimulus condition (.40) were significantly closer to the beginning of the target region than either in the normal unspaced condition (.48), the space before the stimulus condition (.61) or the spaces around the stimulus condition (.57; all p values < .05). Furthermore, initial landing positions in the spaces around the word condition (.59) were significantly away from the word beginning than in the normal unspaced word condition (.48), F<subs>1</subs>(<reflink idref="bib1" id="ref142">1</reflink>, 39) = 5.11, p = .03, F<subs>2</subs>(<reflink idref="bib1" id="ref143">1</reflink>, 79) = 5.98, p = .02. More importantly, a one‐sample t test showed that the initial landing position in the spaces around the word condition (.59) was significantly away from this word centre (.50), t<subs>1</subs>(<reflink idref="bib1" id="ref144">1</reflink>, 39) = 2.24, p = .03, t<subs>1</subs>(<reflink idref="bib1" id="ref145">1</reflink>, 79) = 2.37, p = .02, but the landing position in the normal unspaced word condition (.48) was not significantly away from the word centre (.50), t<subs>1</subs>(<reflink idref="bib1" id="ref146">1</reflink>, 39) = −.77, p = .44; t<subs>1</subs>(<reflink idref="bib1" id="ref147">1</reflink>, 79) = −.93, p = .35. However, there was no significant difference between the spaces around the nonword condition (.55) and the normal unspaced nonword condition (.49; all p values > .20). These results show that inserting spaces around words (thereby facilitating parafoveal word segmentation) did not produce saccades that were aimed at the word centre in order to process the upcoming word in a single fixation.</p> <hd id="AN0129015805-14">Landing position distribution</hd> <p>To provide a more detailed examination of saccade target selection and word segmentation during Chinese reading, we divided the two‐character target region into two single‐character ROIs. Two characters (i.e., AB) near spaces (i.e.,) were analysed in each condition. We referred A to be the character that was next to the space (i.e., A, or A) and B to be the character that was next to A but away from the space (i.e., BA, or AB). In each spacing condition, fixation probability data were subjected to 2 (stimulus: word versus nonword) × 2(character position: first versus second) ANOVAs. As shown in Table  and Figure , there was no significant interaction between stimulus type and character position for each spacing condition (all p values > .20). For both stimuli of word and nonword, the fixation probability on B was significantly higher than that on A (all p values < .01), which indicated that readers were less likely to fixate on the character which was next to the space inserted into the text, regardless of whether the space was before or after this character. In sum, these results showed that saccade targeting was strongly affected by spaces within the text, and these effects were almost identical for stimuli of word and nonword.</p> <p>ANOVA results for fixation probability on single characters.</p> <p> <ephtml> <table border="1" cellpadding="9"><tr><th /><th>Participants analysis</th><th /><th>Items analysis</th><th /></tr><tr><th>Variables</th><th>F1</th><th>MSE</th><th>p</th><th>ηp2</th><th>F2</th><th>MSE</th><th>p</th><th>ηp2</th></tr><tr><td>Normal unspaced condition</td><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td>Stimulus (word versus nonword)</td><td>.00</td><td>156</td><td>.93</td><td>.00</td><td>.00</td><td>234</td><td>.93</td><td>.00</td></tr><tr><td>Character (first versus second)</td><td>.00</td><td>156</td><td>.93</td><td>.00</td><td>.00</td><td>234</td><td>.93</td><td>.00</td></tr><tr><td>Stimulus × character</td><td>.26</td><td>654</td><td>.62</td><td>.00</td><td>.21</td><td>1219</td><td>.65</td><td>.00</td></tr><tr><td>Space before the stimulus condition</td><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td>Stimulus (word versus nonword)</td><td>2.30</td><td>174</td><td>.14</td><td>.06</td><td>2.22</td><td>345</td><td>.14</td><td>.03</td></tr><tr><td>Character (first versus second)</td><td>40.40</td><td>434</td><td><.001</td><td>.51</td><td>56.04</td><td>670</td><td><.001</td><td>.42</td></tr><tr><td>Stimulus × character</td><td>1.37</td><td>257</td><td>.25</td><td>.03</td><td>.95</td><td>681</td><td>.33</td><td>.01</td></tr><tr><td>Space after the stimulus condition</td><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td>Stimulus (word versus nonword)</td><td>18.58</td><td>119</td><td><.001</td><td>.32</td><td>12.10</td><td>288</td><td>.001</td><td>.13</td></tr><tr><td>Character (first versus second)</td><td>12.76</td><td>307</td><td>.001</td><td>.25</td><td>13.29</td><td>710</td><td><.001</td><td>.14</td></tr><tr><td>Stimulus × character</td><td>.12</td><td>284</td><td>.73</td><td>.00</td><td>.07</td><td>835</td><td>.79</td><td>.00</td></tr><tr><td>Spaces around the stimulus condition</td><td /><td /><td /><td /><td /><td /><td /><td /></tr><tr><td>Stimulus (word versus nonword)</td><td>27.86</td><td>98</td><td><.001</td><td>.42</td><td>22.65</td><td>234</td><td><.001</td><td>.22</td></tr><tr><td>Character (first versus second)</td><td>13.47</td><td>520</td><td>.001</td><td>.26</td><td>24.93</td><td>556</td><td><.001</td><td>.24</td></tr><tr><td>Stimulus × character</td><td>.01</td><td>378</td><td>.92</td><td>.000</td><td>.14</td><td>9501</td><td>.71</td><td>.002</td></tr></table> </ephtml> </p> <p>5 Note: For the participant analysis, the degrees of freedom were 1 and 39; and for the items analysis, the degrees of freedom were 1 and 79.</p> <p>Additionally, we found that participants fixated more frequently on characters than spaces (.42 vs .14, all p values < .01) and fixated longer on characters than spaces (250 ms vs 203 ms, all p values < .05) because of information density. In these normal unspaced conditions, there were no significant differences in fixation probability between the first and second characters (.48 and .46 respectively for word stimuli; .46 and .48 respectively for nonword stimuli) or first fixation duration (252 and 253 ms respectively for word stimuli; 253 and 263 ms respectively for nonword stimuli), all p values > .15. As shown in Figure , the preferred viewing location curves were flat in the normal unspaced conditions (for both word and nonword stimuli). These results replicated prior findings that Chinese readers do not select any special position within a word as the saccade target in normal reading (Li et al., [<reflink idref="bib17" id="ref148">17</reflink>] ; Li et al., [<reflink idref="bib18" id="ref149">18</reflink>] ; P. Liu et al., [<reflink idref="bib25" id="ref150">25</reflink>] ; Tsai & McConkie, [<reflink idref="bib42" id="ref151">42</reflink>] ; H. Yang & McConkie, [<reflink idref="bib46" id="ref152">46</reflink>] ).</p> <hd id="AN0129015805-15">General discussion</hd> <p>In order to explain the mechanisms of the different effects of spaces on word recognition in Chinese reading, this study attempted to answer this question: Why would spaces placed after a word help Chinese reader in word recognition? We found that inserting a space after a stimulus facilitated information processing in the word condition but not in the nonword condition. It demonstrates that the facilitatory effect of inserting a space after a word is strongly influenced by word segmentation mechanisms rather than other factors. These results are consistent with the predictions of the word segmentation and recognition model (Li et al., [<reflink idref="bib20" id="ref153">20</reflink>] ), as a space after a word facilitates word segmentation, but a space before this word could not have this positive effect. We also found, for both word and nonword stimuli, Chinese readers tended to fixate on characters away from these spaces inserted into the text due to perceptual economy explanation. These results suggest that Chinese readers' saccade target selection may be strongly associated with current information processing.</p> <p>According to the disparate effects of spaces on fixation durations between words and nonwords, the present study confirmed that the different effects of inserting a space before and after a word are mostly caused by word segmentation mechanisms (P. Liu & Li, [<reflink idref="bib22" id="ref154">22</reflink>] ). It is difficult for Chinese readers to detect where the word ends in sentence reading (Bai et al., [<reflink idref="bib1" id="ref155">1</reflink>] ; Li et al., [<reflink idref="bib20" id="ref156">20</reflink>] ; P. Liu & Li, [<reflink idref="bib22" id="ref157">22</reflink>] ). Inserting a space after word n marks the right boundary of word n, but it does not offer additional information for word n + 1. Thus, a space after word n facilitates word segmentation and recognition for this word, but a space within a word could not have any facilitatory effect for the processing of this word (Li et al., [<reflink idref="bib20" id="ref158">20</reflink>] ). Additionally, if inter‐word spaces cannot be indicators of word boundaries, it is very difficult to explain the finding reported previously. Hence, these results could infer that inter‐word spaces could be an entirely reliable indicator of word boundary location.</p> <p>Importantly, the present study ruled out the possibility that the different effects of inserting a space before and after a word are caused by landing position effects. The results that gaze duration was longer when the eyes fixated on the first character than on the second character within a word in the space after the word condition are completely inconsistent with the OVP effects (P. Liu & Li, [<reflink idref="bib21" id="ref159">21</reflink>] ; P. Liu et al., [<reflink idref="bib23" id="ref160">23</reflink>] ). The present results may be due to information density because there was a space after the second character within a word. The perceptual economy explanation for these results comes about because visual system may learn to expect greater information at the first character than that around the second character. Thus, we found participants fixated more frequently and longer on the first character than on the second character in this space after stimulus condition. The same logic could be used in the space before stimulus condition. Thus, the different effects of spacing locations could not be explained by landing position effects.</p> <p>Additionally, the different effects of spacing locations on word recognition might hardly be explained by parafoveal processing or visual crowding. Compared with the space after stimuli condition, there should be longer fixation times in the space before stimuli condition because the target word was less likely to be processed with parafoveal vision. However, this effect was not seen for gaze duration and first fixation duration of nonword stimuli. Thus, parafoveal processing may not play a key role in the different effects of spacing locations on word recognition. Furthermore, if different effects of spacing locations were indeed strongly influenced by visual crowding rather than word segmentation, we would expect shorter fixation times in the spaces around nonword condition than the space before nonword condition because two spaces would reduce visual crowding effects more. However, there was no result to support it. Thus, these factors of parafoveal processing or visual crowding are not likely to play a key role in the different effects of spacing locations on word recognition.</p> <p>On the other hand, we cannot rule out one possibility that the effects of spaces may be counterbalanced across other factors listed in Table . Compared with normal unspaced text, inserting a space before a word‐disruptive stimulus may produce both positive (i.e., reduction in visual crowding) and negative effects (i.e., reduction in the efficiency of parafoveal processing and disturbing normal reading habits) for processing of this stimulus, and inserting a space after this stimulus might also produce either positive effect (i.e., reduction in visual crowding) or negative effect (i.e., disturbing normal reading habits). The combination of these facilitory and inhibitory effects of spaces for these possible factors may cause the results of similar effects of spaces inserted before and after these word‐disruptive stimuli.</p> <p>There are two components to the issue of eye movement control: where and when to move the eyes. Most studies supported that the decisions of where and when to move the eyes can be made independently in sentence reading (Rayner, [<reflink idref="bib37" id="ref161">37</reflink>] ). Where to move the eyes next is strongly influenced by low‐level information such as word length and space information in English (McGowan et al., [<reflink idref="bib30" id="ref162">30</reflink>] ; McGowan et al., [<reflink idref="bib29" id="ref163">29</reflink>] ; Perea & Acha, [<reflink idref="bib34" id="ref164">34</reflink>] ; Rayner et al., [<reflink idref="bib39" id="ref165">39</reflink>] ; Rayner et al., [<reflink idref="bib40" id="ref166">40</reflink>] ). The present study also indicates that where to move the eyes is influenced by space information. Firstly, initial landing positions were away from the space inserted in Chinese text, regardless of the position in which it was inserted. Second, readers were more likely to fixate on the characters that were away from these spaces compared with these characters next to these spaces, perhaps because of perceptual economy described earlier (for more details, please see P. Liu & Li, [<reflink idref="bib22" id="ref167">22</reflink>] ). Third, although these spaces could not provide any text information, there were some cases where readers' eyes landed on these spaces. These cases may be attributable to the possibility of measurement error or mislocated fixations, as saccades occasionally do not land on their intended targets (McConkie, Kerr, Reddix & Zola, [<reflink idref="bib28" id="ref168">28</reflink>] ; Nuthmann, Engbert & Kliegl, [<reflink idref="bib32" id="ref169">32</reflink>] ). Interestingly, the three patterns described previously were almost identical for word and nonword stimuli, and this could be interpreted as the oculomotor effects that had nothing to do with fixation durations of word identification.</p> <p>Our results challenge a hypothesis that readers' saccade targeting is influenced by parafoveal word segmentation in Chinese (Yan et al., [<reflink idref="bib45" id="ref170">45</reflink>] ). The hypothesis states that readers would move to the centre of the upcoming word in a single fixation if they successfully segment the word in the parafovea. However, if readers fail to segment the word within parafoveal vision, they would initially move to the beginning of the upcoming word and refixate it. According to this hypothesis, if Chinese readers use these spaces as word segmentation cues, it would be expected that conditions that would facilitate parafoveal word segmentation would result in saccades being targeted at the words' centre to process the upcoming word in a single fixation. However, these effects were not found. Firstly, we observed that initial landing positions for single fixations were significantly further from the word centre in the spaces around the word condition than in the normal unspaced condition. Secondly, we observed that landing positions in single fixation cases were almost identical in the spaced nonword conditions to those in the spaced word conditions. Finally, consistent with the prior results (Li et al., [<reflink idref="bib17" id="ref171">17</reflink>] ; Li et al., [<reflink idref="bib18" id="ref172">18</reflink>] ; P. Liu et al., [<reflink idref="bib25" id="ref173">25</reflink>] ; Tsai & McConkie, [<reflink idref="bib42" id="ref174">42</reflink>] ; H. Yang & McConkie, [<reflink idref="bib46" id="ref175">46</reflink>] ), we did not find that Chinese readers selected any special position within a word as the saccade target in the normal unspaced condition. Thus, these results indicate that saccade target selection may be not strongly determined by parafoveal word segmentation in these spacing conditions.</p> <p>Although the present study can shed some light on the issue of word segmentation and eye movement control in Chinese reading, there were some limitations to the spacing presentations used. During normal reading of Chinese, there are no visual cues such as spaces to mark word boundaries between written words. Despite the visually unfamiliar layout of spaced Chinese text, previous studies have found that inserting spaces between words does not interfere with normal reading (Bai et al., [<reflink idref="bib1" id="ref176">1</reflink>] ; Hsu & Huang, [<reflink idref="bib10" id="ref177">10</reflink>] , [<reflink idref="bib11" id="ref178">11</reflink>] ; I. M. Liu et al., [<reflink idref="bib26" id="ref179">26</reflink>] ; P. Liu & Li, [<reflink idref="bib22" id="ref180">22</reflink>] ; Zang et al., [<reflink idref="bib48" id="ref181">48</reflink>] ). Indeed, our results also indicated that spaces after word conditions provide useful word boundary information, and Chinese readers may regard these spaces as reliable cues for word boundaries. The actual effect of inserting spaces in Chinese text remains unclear, and the present study is a step forward to clarify the origin of the space effects. For instance, placing the spaces at different positions of sentence structure (e.g., subject‐verb structure) may produce different impacts, and these issues need further research.[<reflink idref="bib8" id="ref182">8</reflink>] Additionally, the present study cannot rule out the possibility that the effects of low‐level information such as spaces on eye movements were so robust that other effects were neglected. Thus, participants might select saccade targets based on other strategies rather than parafoveal word segmentation.[<reflink idref="bib9" id="ref183">9</reflink>] One possible further research is still needed to explore the issue of which factors determine where to move the eyes in Chinese reading.</p> <p>In summary, our results indicate that the different effects of inserting a space before and after a word are mainly attributable to word segmentation mechanisms, which is consistent with the prediction of a word segmentation and word recognition model of Chinese reading. Finally, the results demonstrate that Chinese readers' oculomotor control systems do show flexibility in that they can adapt to unusual presentation formats quickly based on perceptual economic efficiency.</p> <hd id="AN0129015805-16">Acknowledgements</hd> <p>This research was partially supported by the National Natural Science Foundation of China (No. 31070904, 31600887), the Hong Kong Scholars Program (No. XJ2015050) and by the CAS Key Laboratory of Mental Health (No. KLMH2014ZG14). We thank Xingshan Li, Ming Yan and Victoria A. McGowan for their helpful input and comments on an earlier version of this article. We also thank Lin Gui for his helpful comments concerning statistical issues addressed in this article.</p> <ref id="AN0129015805-17"> <title>Notes</title> <blist> <bibl id="bib6" idref="ref9" type="bt">6</bibl> <bibtext>Under some circumstances (e.g., ambiguous sentences or for naïve readers), inter‐word spaces or other segmentation cues can play important roles in sentence comprehension, partly because they may help distinguish between alternative meanings (Blythe et al., ; Hsu & Huang, , ). </bibtext> </blist> <blist> <bibl id="bib7" idref="ref12" type="bt">7</bibl> <bibtext>It was not possible to report data for initial landing position for the first fixation in multiple fixation cases because of a lack of data points. </bibtext> </blist> <blist> <bibl id="bib8" idref="ref42" type="bt">8</bibl> <bibtext>We are grateful to an anonymous reviewer for raising this issue and the one that has just been discussed. </bibtext> </blist> <blist> <bibl id="bib9" idref="ref43" type="bt">9</bibl> <bibtext>We are grateful to Dr Ming Yan for raising this issue and the one discussed previously. </bibtext> </blist> </ref> <ref id="AN0129015805-18"> <title>References</title> <blist> <bibl id="bib1" idref="ref10" type="bt">1</bibl> <bibtext>Bai, X., Yan, G., Liversedge, S.P., Zang, C. & Rayner, K. (2008). Reading spaced and unspaced Chinese text: Evidence from eye movements. Journal of Experimental Psychology: Human Perception and Performance, 34, 1277–1287. DOI:10.1037/0096‐1523.34.5.1277. </bibtext> </blist> <blist> <bibl id="bib2" idref="ref11" type="bt">2</bibl> <bibtext>Blythe, H.I., Liang, F., Zang, C., Wang, J., Yan, G., Bai, X. et al. (2012). Inserting spaces into Chinese text helps readers to learn new words: An eye movement study. Journal of Memory and Language, 67, 241–254. DOI:10.1016/j.jml.2012.05.004. </bibtext> </blist> <blist> <bibl id="bib3" idref="ref23" type="bt">3</bibl> <bibtext>Bouma, H. (1970). Interaction effects in parafoveal letter recognition. Nature, 226, 177–178. DOI:10.1038/226177a0. </bibtext> </blist> <blist> <bibl id="bib4" idref="ref24" type="bt">4</bibl> <bibtext>Bouma, H. (1973). Visual interference in the parafoveal recognition of initial and final letters of words. Vision Research, 13, 767–782. DOI:10.1016/0042‐6989(73)90041‐2. </bibtext> </blist> <blist> <bibl id="bib5" idref="ref45" type="bt">5</bibl> <bibtext>Bricolo, E., Salvi, C., Martelli, M., Arduino, L.S. & Daini, R. (2015). The effects of crowding on eye movement patterns in reading. Acta Psychologica, 160, 23–34. DOI:10.1016/j.actpsy.2015.06.003. </bibtext> </blist> <blist> <bibtext>Brysbaert, M. & Nazir, T.A. (2005). Visual constraints in written word recognition: Evidence from the optimal viewing‐position effect. Journal of Research in Reading, 28, 216–228. DOI:10.1111/j.1467‐9817.2005.00266.x. </bibtext> </blist> <blist> <bibtext>Cui, L., Drieghe, D., Bai, X., Yan, G. & Liversedge, S.P. (2014). Parafoveal preview benefit in unspaced and spaced Chinese reading. The Quarterly Journal of Experimental Psychology, 67, 2172–2188. DOI:10.1080/17470218.2014.909858. </bibtext> </blist> <blist> <bibtext>Cui, L., Drieghe, D., Yan, G.L., Bai, X.J., Chi, H. & Liversedge, S.P. (2013). Parafoveal processing across different lexical constituents in Chinese reading. The Quarterly Journal of Experimental Psychology, 66, 403–416. DOI:10.1080/17470218.2012.720265. </bibtext> </blist> <blist> <bibtext>Drieghe, D. (2011). Parafoveal‐on‐foveal effects on eye movements during reading. In S.P. Liversedge, I.D. Gilchrist & S. Everling (Eds.), The Oxford handbook on eye movements, (pp. 839–855). Oxford, UK: Oxford University Press. </bibtext> </blist> <blist> <bibl id="bib10" idref="ref177" type="bt">10</bibl> <bibtext>Hsu, S. & Huang, K. (2000a). Effects of word spacing on reading Chinese text from a video display terminal. Perceptual and Motor Skills, 90, 81–92. DOI:10.2466/pms.2000.90.1.81. </bibtext> </blist> <blist> <bibl id="bib11" idref="ref178" type="bt">11</bibl> <bibtext>Hsu, S. & Huang, K. (2000b). Interword spacing in Chinese text layout. Perceptual and Motor Skills, 91, 355–365. DOI:10.2466/pms.2000.91.2.355. </bibtext> </blist> <blist> <bibl id="bib12" idref="ref22" type="bt">12</bibl> <bibtext>Inhoff, A.W. & Wu, C. (2005). Eye movements and the identification of spatially ambiguous words during Chinese sentence reading. Memory & Cognition, 33, 1345–1356. DOI:10.3758/BF03193367. </bibtext> </blist> <blist> <bibl id="bib13" idref="ref13" type="bt">13</bibl> <bibtext>Inhoff, A.W., Liu, W., Wang, J. & Fu, D.J. (1997). Use of spatial information during the reading of Chinese text. In D.L. Peng, H. Shu & H.C. Chen (Eds.), Cognitive research on Chinese language, (pp. 296–329). Jinan, China: Shan Dong Educational Publishing. </bibtext> </blist> <blist> <bibl id="bib14" idref="ref44" type="bt">14</bibl> <bibtext>Kennedy, A. & Pynte, J. (2005). Parafoveal‐on‐foveal effects in normal reading. Vision Research, 45, 153–168. DOI:10.1016/j.visres.2004.07.037. </bibtext> </blist> <blist> <bibl id="bib15" idref="ref39" type="bt">15</bibl> <bibtext>Kennison, S.M. & Clifton, C. (1995). Determinants of parafoveal preview benefit in high and low working memory capacity readers: Implications for eye movement control. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 68–81. </bibtext> </blist> <blist> <bibl id="bib16" idref="ref64" type="bt">16</bibl> <bibtext>Chinese Lexicon. (2003). [CD‐ROM]. Produced by State Key Laboratory of Intelligent Technology and Systems, Tsinghua University and Institute of Automation, Chinese Academy of Sciences. Retrieved from Chinese linguistic Data Consortium (<ulink href="http://www.chineseldc.org/">http://www.chineseldc.org/</ulink>). </bibtext> </blist> <blist> <bibl id="bib17" idref="ref52" type="bt">17</bibl> <bibtext>Li, X., Bicknell, K., Liu, P., Wei, W. & Rayner, K. (2014). Reading is fundamentally similar across disparate writing systems: A systematic characterization of how words and characters influence eye movements in Chinese reading. Journal of Experiment Psychology: General, 143, 895–913. DOI:10.1037/a0033580. </bibtext> </blist> <blist> <bibl id="bib18" idref="ref53" type="bt">18</bibl> <bibtext>Li, X., Liu, P. & Rayner, K. (2011). Eye movement guidance in Chinese reading: Is there a preferred viewing location? Vision Research, 51, 1146–1156. DOI:10.1016/j.visres.2011.03.004. </bibtext> </blist> <blist> <bibl id="bib19" idref="ref54" type="bt">19</bibl> <bibtext>Li, X., Liu, P. & Rayner, K. (2015). Saccade target selection in Chinese reading. Psychonomic Bulletin & Review, 22, 524–530. DOI:10.3758/s13423‐014‐0693‐3. </bibtext> </blist> <blist> <bibl id="bib20" idref="ref33" type="bt">20</bibl> <bibtext>Li, X., Rayner, K. & Cave, K.R. (2009). On the segmentation of Chinese words during reading. Cognitive Psychology, 58, 525–552. DOI:10.1016/j.cogpsych.2009.02.003. </bibtext> </blist> <blist> <bibl id="bib21" idref="ref37" type="bt">21</bibl> <bibtext>Liu, P. & Li, X. (2013). Optimal viewing position effects in the processing of isolated Chinese words. Vision Research, 81, 45–57. DOI:10.1016/j.visres.2013.02.004. </bibtext> </blist> <blist> <bibl id="bib22" idref="ref18" type="bt">22</bibl> <bibtext>Liu, P. & Li, X. (2014). Inserting spaces before and after words affect word processing differently in Chinese: Evidence from eye movements. British Journal of Psychology, 105, 57–68. DOI:10.1111/bjop.12013. </bibtext> </blist> <blist> <bibl id="bib23" idref="ref38" type="bt">23</bibl> <bibtext>Liu, P., Liu, D., Han, B. & Paterson, K.B. (2015). Aging and optimal viewing position effect in Chinese. Frontiers in Psychology, 6, 1–13. DOI:10.3389/fpsyg.2015.01656. </bibtext> </blist> <blist> <bibl id="bib24" idref="ref56" type="bt">24</bibl> <bibtext>Liu, Y., Reichle, E.D. & Li, X. (2015). Parafoveal processing affects outgoing saccade length during the reading of Chinese. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41, 1229–1236. DOI:10.1037/xlm0000057. </bibtext> </blist> <blist> <bibl id="bib25" idref="ref55" type="bt">25</bibl> <bibtext>Liu, P., Li, W., Han, B. & Li, X. (2014). Effects of anomalous characters and small stroke omissions on eye movements during the reading of Chinese sentences. Ergonomics, 57, 1659–1669. DOI:10.1080/00140139.2014.945492. </bibtext> </blist> <blist> <bibl id="bib26" idref="ref14" type="bt">26</bibl> <bibtext>Liu, I.M., Yeh, J.S., Wang, L.H. & Chang, Y.K. (1974). Effects of arranging Chinese words as units on reading efficiency. Chinese Journal of Psychology, 16, 25–32. </bibtext> </blist> <blist> <bibl id="bib27" idref="ref57" type="bt">27</bibl> <bibtext>Ma, G., Li, X. & Pollatsek, A. (2015). There is no relationship between preferred viewing location and word segmentation in Chinese reading. Visual Cognition, 23, 399–414. DOI:10.1080/13506285.2014.1002554. </bibtext> </blist> <blist> <bibl id="bib28" idref="ref168" type="bt">28</bibl> <bibtext>McConkie, G.W., Kerr, P.W., Reddix, M.D. & Zola, D. (1988). Eye movement control during reading: I. The location of initial eye fixations on words. Vision Research, 28, 1107–1118. </bibtext> </blist> <blist> <bibl id="bib29" idref="ref1" type="bt">29</bibl> <bibtext>McGowan, V.A., White, S.W., Jordan, T.R. & Paterson, K.B. (2014). Aging and the use of inter‐word spaces during reading: Evidence from eye movements. Psychonomic Bulletin & Review, 21, 740–747. DOI:10.3758/s13423‐013‐0527‐8. </bibtext> </blist> <blist> <bibl id="bib30" idref="ref26" type="bt">30</bibl> <bibtext>McGowan, V.A., White, S.J. & Paterson, K.B. (2015). The effects of interword spacing on the eye movements of young and older readers. Journal of Cognitive Psychology, 27, 609–621. DOI:10.1080/20445911.2014.988157. </bibtext> </blist> <blist> <bibl id="bib31" idref="ref2" type="bt">31</bibl> <bibtext>Morris, R.K., Rayner, K. & Pollatsek, A. (1990). Eye movement guidance in reading: The role of parafoveal letter and space information. Journal of Experimental Psychology: Human Perception and Performance, 16, 268–281. DOI:10.1037/0096‐1523.16.2.268. </bibtext> </blist> <blist> <bibl id="bib32" idref="ref169" type="bt">32</bibl> <bibtext>Nuthmann, A., Engbert, R. & Kliegl, R. (2005). Mislocated fixations during reading and the inverted optimal viewing position effect. Vision Research, 45, 2201–2217. DOI:10.1016/j.visres.2005.02.014. </bibtext> </blist> <blist> <bibl id="bib33" idref="ref58" type="bt">33</bibl> <bibtext>Pan, J., Yan, M., Laubrock, J., Shu, H. & Kliegl, R. (2014). Saccade‐target selection of dyslexic children when reading Chinese. Vision Research, 97, 24–30. DOI:10.1016/j.visres.2014.01.014. </bibtext> </blist> <blist> <bibl id="bib34" idref="ref3" type="bt">34</bibl> <bibtext>Perea, M. & Acha, J. (2009). Space information is important for reading. Vision Research, 49, 1994–2000. DOI:10.1016/j.visres.2009.05.009. </bibtext> </blist> <blist> <bibl id="bib35" idref="ref29" type="bt">35</bibl> <bibtext>Perea, M., Moret‐Tatay, C. & Gomez, P. (2011). The effects of interletter spacing in visual‐word recognition. Acta Psychologica, 137, 345–351. DOI:10.1016/j.actpsy.2011.04.003. </bibtext> </blist> <blist> <bibl id="bib36" idref="ref4" type="bt">36</bibl> <bibtext>Pollatsek, A. & Rayner, K. (1982). Eye movement control in reading: The role of word boundaries. Journal of Experimental Psychology: Human Perception and Performance, 8, 817–833. DOI:10.1037/0096‐1523.8.6.817. </bibtext> </blist> <blist> <bibl id="bib37" idref="ref40" type="bt">37</bibl> <bibtext>Rayner, K. (2009). Eye movements and attention in reading, scene perception, and visual search. The Quarterly Journal of Experimental Psychology, 62, 1457–1506. DOI:10.1080/17470210902816461. </bibtext> </blist> <blist> <bibl id="bib38" idref="ref5" type="bt">38</bibl> <bibtext>Rayner, K., Fischer, M.H. & Pollatsek, A. (1998). Unspaced text interferes with both word identification and eye movement control. Vision Research, 38, 1129–1144. DOI:10.1016/S0042‐6989(97)00274‐5. </bibtext> </blist> <blist> <bibl id="bib39" idref="ref6" type="bt">39</bibl> <bibtext>Rayner, K., Slattery, T.J. & Belanger, N.N. (2010). Eye movements, the perceptual span, and reading speed. Psychonomic Bulletin & Review, 17, 834–839. DOI:10.3758/pbr.17.6.834. </bibtext> </blist> <blist> <bibl id="bib40" idref="ref7" type="bt">40</bibl> <bibtext>Rayner, K., Yang, J., Schuett, S. & Slattery, T. (2013). Eye movements of older and younger readers when reading unspaced text. Experimental Psychology, 60, 354–361. DOI:10.1027/1618‐3169/a000207. </bibtext> </blist> <blist> <bibl id="bib41" idref="ref59" type="bt">41</bibl> <bibtext>Shu, H., Zhou, W., Yan, M. & Kliegl, R. (2011). Font size modulates saccade‐target selection in Chinese reading. Attention, Perception & Psychophysics, 73, 482–490. DOI:10.3758/s13414‐010‐0029‐y. </bibtext> </blist> <blist> <bibl id="bib42" idref="ref151" type="bt">42</bibl> <bibtext>Tsai, J.L. & McConkie, G.W. (2003). Where do Chinese readers send their eyes? In J. Hyona, R. Radach & H. Deubel (Eds.), The mind's eye: Cognitive and applied aspects of eye movement research, (pp. 159–176). Oxford: Elsevier Science. </bibtext> </blist> <blist> <bibl id="bib43" idref="ref66" type="bt">43</bibl> <bibtext>von der Malsburg, T., & Angele, B. (2015). False positive rates in standard analyses of eye movements in reading. <ulink href="http://arxiv.org/abs/1504.06896,">http://arxiv.org/abs/1504.06896,</ulink> 1–18 </bibtext> </blist> <blist> <bibl id="bib44" idref="ref8" type="bt">44</bibl> <bibtext>Winskel, H., Radach, R. & Luksanneeyanawin, S. (2009). Eye movements when reading spaced and unspaced Thai and English: A comparison of Thai‐English bilinguals and English monolinguals. Journal of Memory and Language, 61, 339–351. DOI:10.1016/j.jml.2009.07.002. </bibtext> </blist> <blist> <bibl id="bib45" idref="ref60" type="bt">45</bibl> <bibtext>Yan, M., Kliegl, R., Richter, E.M., Nuthmann, A. & Shu, H. (2010). Flexible saccade‐target selection in Chinese reading. The Quarterly Journal of Experimental Psychology, 63, 705–725. DOI:10.1080/17470210903114858. </bibtext> </blist> <blist> <bibl id="bib46" idref="ref152" type="bt">46</bibl> <bibtext>Yang, H. & McConkie, G.W. (1999). Reading Chinese: Some basic eye‐movement characteristics. In J. Wang, A.W. Inhoff & H. Chen (Eds.), Reading Chinese script, (pp. 207–222). Mahwah, NJ: Lawrence Erlbaum Associates. </bibtext> </blist> <blist> <bibl id="bib47" idref="ref41" type="bt">47</bibl> <bibtext>Yang, J., Li, N., Wang, S., Slattery, T. & Rayner, K. (2014). Encoding the target or the plausible preview word? The nature of the plausibility preview benefit in reading Chinese. Visual Cognition, 22, 193–213. DOI:10.1080/13506285.2014.890689. </bibtext> </blist> <blist> <bibl id="bib48" idref="ref15" type="bt">48</bibl> <bibtext>Zang, C., Liang, F., Bai, X., Yan, G. & Liversedge, S.P. (2013). Interword spacing and landing position effects during Chinese reading in children and adults. Journal of Experimental Psychology: Human Perception and Performance, 39, 720–734. DOI:10.1037/a0030097. </bibtext> </blist> <blist> <bibl id="bib49" idref="ref65" type="bt">49</bibl> <bibtext>Zhou, Y. (1980). Hanzi Shengpang Duyin Biancha (A Handy lookup for the pronunciation of phonetics in Chinese characters). Jilin: Jilin People's Press. </bibtext> </blist> </ref> <p>PHOTO (COLOR): Examples of word boundaries English translation of this sentence: This charity organisation has been appealing to the public to pay attention to children with mental retardation for many years.</p> <p>PHOTO (COLOR): Means and standard errors for first fixation durations on the two‐character region of interest across the different spacing conditions.</p> <p>PHOTO (COLOR): Means and standard errors for gaze durations on the two‐character region of interest across the different spacing conditions.</p> <p>PHOTO (COLOR): Means and standard errors for total times on the two‐character region of interest across the different spacing conditions.</p> <p>PHOTO (COLOR): Means and standard errors for initial landing positions on the two‐character region of interest across the different spacing conditions.</p> <p>PHOTO (COLOR): Means and standard errors for initial landing positions for single fixations on the two‐character region of interest across the different spacing conditions.</p> <p>PHOTO (COLOR): Proportion of initial fixations on the first‐character and second‐character region of interests across the different presentation conditions.</p> <aug> <p>By Pingping Liu and Qin Lu</p> <p></p> <p>Pingping Liu is an associate professor from the Institute of Psychology, Chinese Academy of Sciences. Currently, she is a postdoctoral fellow at the Hong Kong Polytechnic University. Dr Liu's research includes word recognition, eye movements during reading, and individual differences in reading across the life‐span.</p> <p>Qin Lu is a full professor in the Department of Computing at the Hong Kong Polytechnic University. Prof. Lu's research includes natural language process and computational linguistics, lexical semantics, information extraction and knowledge discovery, and open systems and standardization.</p> </aug> <nolink nlid="nl1" bibid="bib751" firstref="ref107"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1175754
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: The Effects of Spaces on Word Segmentation in Chinese Reading: Evidence from Eye Movements
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Liu%2C+Pingping%22">Liu, Pingping</searchLink><br /><searchLink fieldCode="AR" term="%22Lu%2C+Qin%22">Lu, Qin</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Journal+of+Research+in+Reading%22"><i>Journal of Research in Reading</i></searchLink>. May 2018 41(2):329-349.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley-Blackwell. 350 Main Street, Malden, MA 02148. Tel: 800-835-6770; Tel: 781-388-8598; Fax: 781-388-8232; e-mail: cs-journals@wiley.com; Web site: http://www.wiley.com/WileyCDA
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 21
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2018
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Eye+Movements%22">Eye Movements</searchLink><br /><searchLink fieldCode="DE" term="%22Chinese%22">Chinese</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Processes%22">Reading Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Orthographic+Symbols%22">Orthographic Symbols</searchLink><br /><searchLink fieldCode="DE" term="%22Language+Processing%22">Language Processing</searchLink><br /><searchLink fieldCode="DE" term="%22Phonemes%22">Phonemes</searchLink><br /><searchLink fieldCode="DE" term="%22Reading+Research%22">Reading Research</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/1467-9817.12106
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0141-0423
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: This paper studies the mechanisms behind the differential effects of inserting a space either before or after a two-character unit on information processing through the examination of eye movements in Chinese, a language where there is no word delimiters. A two-character unit in this study is either a word-preserving stimulus or a word-disrupting stimulus (i.e., nonword). The study aims for a better understanding of the cognitive mechanisms for lexical processing that may underlie observed facilitory or inhibitory effects of the spacing conditions in sentence context. Results show that inserting a space after a word facilitates lexical processing, but inserting a space before a word does not. Inserting a space before and after a nonword, however, does not show these effects. These results indicate that the effects of spaces before and after words are mainly influenced by word segmentation mechanisms rather than landing position effects or other factors.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2018
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1175754
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1175754
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/1467-9817.12106
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 21
        StartPage: 329
    Subjects:
      – SubjectFull: Eye Movements
        Type: general
      – SubjectFull: Chinese
        Type: general
      – SubjectFull: Reading Processes
        Type: general
      – SubjectFull: Orthographic Symbols
        Type: general
      – SubjectFull: Language Processing
        Type: general
      – SubjectFull: Phonemes
        Type: general
      – SubjectFull: Reading Research
        Type: general
    Titles:
      – TitleFull: The Effects of Spaces on Word Segmentation in Chinese Reading: Evidence from Eye Movements
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Liu, Pingping
      – PersonEntity:
          Name:
            NameFull: Lu, Qin
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 05
              Type: published
              Y: 2018
          Identifiers:
            – Type: issn-print
              Value: 0141-0423
          Numbering:
            – Type: volume
              Value: 41
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Journal of Research in Reading
              Type: main
ResultId 1