The effects of Chinese and English background music on Chinese reading comprehension

Abstract

The study investigated the influence of the language of background music with lyrics (Chinese and English) on Chinese reading comprehension after controlling for English listening proficiency. A total of 38 Chinese English learners from a university were enrolled in a within-subject design experiment. All participants were exposed to three background conditions: silence, English background music with lyrics (EL-BGM), and Chinese background music with lyrics (CL-BGM). The results showed that participants’ Chinese reading performance was negatively influenced by EL-BGM, with significantly lower reading efficiency and reading rate. Participants with better English listening ability had higher Chinese reading efficiency under the EL-BGM condition. Possible explanations for these findings in terms of the interference-by-process hypothesis, the attentional capture account, and the duplex-mechanism account are discussed.

Keywords

background lyrical music reading comprehension irrelevant sound effect

The irrelevant sound effect (ISE), referring to the primary task performance that can be impaired by a to-be-ignored background sound (Perham & Currie, 2014), is commonly researched. Over the past 80 years, many researchers have conducted experiments to examine the effect on a complex cognitive task, reading.

Concerning the impact of task-irrelevant sound on reading, there are two representative theoretical accounts. The phonological-interference hypothesis considers that the irrelevant sound automatically accesses the phonological loop, coexists, and corrupts the English reading phonological representations (Hyönä & Ekholm, 2016). Based on this theory, the ISE does not depend on the meaningfulness of the speech. On the contrary, the interference-by-process hypothesis states that the interference results from both the irrelevant sound and primary task competing for the same processing mechanism (Anderson & Fuller, 2010). In this case, understanding the meaning of reading is impaired by a meaningful unrelated sound, which goes through the same semantic processing.

Recently, Vasilev et al. (2018) conducted a meta-analysis study, showing that background music had a reliably negative effect on reading comprehension, and music with lyrics was significantly more distracting than instrumental music because lyrics in participants’ mother tongue are intelligible to them. This result supported the interference-by-process hypothesis which emphasizes the meaningfulness of irrelevant sound, while rejecting the phonological-interference hypothesis.

There are an increasing number of people who understand more than one language. However, most studies have examined the effect when both the language of the lyrics and the readings are in participants’ native language (Anderson & Fuller, 2010; Avila et al., 2011; Perham & Currie, 2014). Foreign language learners’ foreign language proficiency is scarcely explored. However, Chew et al. (2016) and Hyönä and Ekholm (2016) studied the effect of the language of background music lyrics on reading performance; the foreign language of background music was meaningless and unintelligible to participants, which cannot demonstrate the effect of language proficiency.

Therefore, the present study intends to study: (1) the effect of the language of background music (English and Chinese) on Chinese reading comprehension performance in Chinese English learners after controlling their English listening proficiency and (2) the relationship between English listening proficiency and reading comprehension performance in the English background music condition. Predicted by the interference-by-process hypothesis, both Chinese and English background music should impair reading performance because participants can understand and process both languages semantically, which compete for the semantic processing that reading needs. More proficient English learners may be more distracted by English background music as they are better at semantically processing the lyrics (Vasilev et al., 2018).

Method

Participants

Expecting to observe a similar effect to related studies (Ren & Xu, 2019), this study used the G*Power 3.1 software package (Faul et al., 2009) for repeated measures (within factors) analysis of variance (ANOVA) to reach the statistical power, parameters: effect size specification as in SPSS, $η_{p}^{2}$ = 0.13, f(U) = 0.39, α = .05, power = 0.80, number of groups = 1, number of measurements = 3, non-sphericity correction = 1; total calculated sample size = 35. A total of thirty-nine 20- to 22-year-old undergraduate Chinese students (M = 21.6, SD = 0.64) participated in this study (34 females). They were from a university that uses English as a medium of instruction in Zhuhai, China for at least 2.5 years. One of the participants had a reaction time that was much lower than three standard deviations below the means across the three conditions, so that, his data were removed from the following analysis.

Compliance with ethical standards

The university where this research was conducted did not have an ethics committee. Nevertheless, this study was approved by the department of psychology. All individual adult participants signed an informed consent form before taking part in the experiment.

Materials

Reading materials

Three Chinese reading articles (1,027, 1,034, and 867 words, respectively) were chosen from the Chinese version of “An Introduction to Special Education” (Hallahan et al., 2010). Five true-or-false questions were created for each article by the researcher. The statistical difficulties¹ of each article’s comprehension questions were .85, .83, and .80. There was also no statistical difference in accuracy and subjective difficulty ratings among the three articles, which provided further evidence that these three articles were equivalent in difficulty.

Background music conditions

There were three background sound conditions: (1) silence, (2) music with lyrics in English, and (3) music with lyrics in Chinese. To reduce variability (e.g., musical styles, rhythms, dissonance, tonality), the song “Let It Go”² was used with the same melody in different languages in (2) and (3). Their tempi and loudness were adjusted to the same level by the “AudioRetoucher” (AbyssMedia, 2016) and “Wave Editor” (AbyssMedia, 2014) software.

English listening ability

An English listening test with 10 questions was retrieved from one Listening Section of the Cambridge IELTS 7 (Cambridge ESOL, 2009) to assess participants’ English listening abilities (ELAs). The following ELA refers to accuracy in the listening test. In this study, participants’ ELA ranged from 1 to 10 (M = 5.05).

Procedure

This study was carried out in a quiet computer laboratory with several cubicles. Each cubicle had a headset and a desktop computer that had installed the E-Prime 2.0 (Schneider et al., 2001). Participants conducted the experiment in separate cubicles individually. The study was a within-subject design experiment. Each participant read three articles in three background music conditions, including silence, English background music with lyrics (EL-BGM), and Chinese background music with lyrics (CL-BGM). E-prime 2.0 was used to present instructions, ratings, music, and questions. The readings were printed on pieces of one-sided A4 paper, with one reading on one page. Background music excerpts were presented through the headset.

The order of presenting the articles was fixed. The presented order of the three conditions was counterbalanced by a 3 × 3 Latin square. Participants were randomly assigned to one of the three groups with different presenting orders of background music. Three one-way ANOVAs showed that there was no order effect among the three groups across the three background conditions: silence—F(1, 35) = .23, p = .79; EL-BGM—F(1, 35) = .97, p = .39; CL-BGM—F(1, 35) = .42, p = .66.

The general procedure is shown in Figure 1. First, the participants read the instructions and completed the English listening test. Then, the experiment with three cycles (three background music conditions) started. During the reading phase, after reading the article with or without background music, they pressed the “Y” button in the E-Studio to indicate that they had finishing reading (E-Prime recorded reaction time). After that, they rated the subjective difficulty of the article on a 5-point Likert-type scale. Then, five true-or-false questions showed up one after another (E-Prime recorded accuracy). Before beginning in another cycle, they were forced to watch a 4-min cartoon to alleviate fatigue and practice effects.

Figure 1.

The General Layout of the Experiment Procedure.

Data analysis

Statistical analysis was performed using IBM SPSS 26.0. Vasilev et al. (2018) pointed out that previous studies mostly paid attention to the accuracy of reading comprehension, whereas a few focused on reading speed and reading efficiency. Therefore, this study computed three reading measures for each participant: accuracy (the number of right answers/total number of questions), reading rate (the number of words in the article/reaction time), and reading efficiency (reading rate multiplied by accuracy; Kallinen, 2002).

The Kolmogorov–Smirnov test justified that reading rate and reading efficiency across conditions were normally distributed, while accuracy and ELA were not. Thus, the analyses of reading rate and reading efficiency employed parametric tests. In contrast, statistical analyses involving accuracy and ELA were analyzed through non-parametric tests.

Results

Table 1 shows the results of reading measures in the three background music conditions (silence, EL-BGM, and CL-BGM).

Table 1.

Means (With SD) of Reading Measures in Three Background Music Conditions.

Background conditions	Reading rate	Accuracy	Reading efficiency
Silence	6.75 (3.06)	0.84 (0.15)	5.51 (2.15)
EL-BGM	5.90 (2.41)	0.78 (0.23)	4.47 (1.92)
CL-BGM	6.32 (2.86)	0.86 (0.17)	5.25 (2.28)

Note. N = 38. EL-BGM = English background music with lyrics; CL-BGM = Chinese background music with lyrics.

To assess the first research question for reading rate and efficiency, two repeated-measured analysis of covariance (ANCOVAs) with mean-centered ELA were performed to estimate the within x covariate interaction (Schneider et al., 2015). The results indicated that the interaction between background music conditions and ELA did not approach significance when the dependent variable was reading efficiency (p = .918) or reading rate (p = .507). This implies that the effect of background music did not change across the range of ELA. Next, a standard repeated measure ANOVA was employed to test the main effect of music condition, because the covariate may contaminate the mean square for the main effect in an ANCOVA (Schneider et al., 2015). There was a significant main effect on reading efficiency, F(2, 74) = 4.502, MSE = 2.498, p = .014, $η_{p}^{2}$ = .108. Three paired samples t-tests revealed significant differences between EL-BGM and silence (p = .008), between EL-BGM and CL-BGM (p = .014), but not between silence and CL-BGM (p = .528).

For analyzing reading rate by repeated measures ANOVA, because it violated the Mauchly’s sphericity test, a Greenhouse–Geisser correction was used to adjust for the lack of sphericity. A marginally significant effect was found, F(2, 74) = 2.961, MSE = 2.710, p = .066, and $η_{p}^{2}$ = .074. Three paired samples t-tests were employed to analyze the statistical difference among the three conditions. The difference between silence and EL-BGM was significant, t(1, 37) = 2.234, p = .032, but there were no significant differences between the other two conditions.

Because the data of ELA and accuracy across three conditions did not fit the normal distribution, non-parametric statistical tests were used to analyze these two variables. A Friedman test showed that there was no difference in accuracy among the three conditions (p = .138). Nevertheless, the means revealed a negative trend of EL-BGM, which had lower accuracy compared with the other two conditions. To answer the second research question, a Spearman correlation indicated a marginally significant positive relationship between ELA and reading efficiency in EL-BGM (ρ = .322, p = .049). No relationship with other variables (e.g., reading efficiency in CL-BGM, reading rate in EL-BGM) was found.

Discussion

The objective of the current study was to assess the effect of the language of intelligible background music with lyrics on Chinese reading comprehension, to reveal the role of language proficiency in ISE. The results showed that, after controlling for Chinese English learners’ ELA, their Chinese reading performance was negatively influenced by EL-BGM, with significantly lower reading efficiency and reading rate. No significant difference between silence and CL-BGM was observed. The correlation between ELA and reading efficiency under the EL-BGM condition was positive and statistically significant.

Focusing on the accuracy of reading, this study found that Chinese reading comprehension was affected by neither Chinese nor English background music with lyrics. This result was inconsistent with the meta-analysis findings of Vasilev et al. (2018), which showed the distracting effect of intelligible background music with lyrics on readings. What is noteworthy is that most of the studies that they analyzed employed English readings, but there is a linguistic difference between processing English and Chinese readings. When reading English, a phonetic language, the texts are converted to subvocal speech, then these phonological representations were retained and stored in the phonological loop system (Hyönä & Ekholm, 2016; Martin et al., 1988). By contrast, while reading Chinese, an ideographic language, the materials can be transformed from visual into semantic information directly, which depends less on the phonological mechanism (Ren & Xu, 2019). As task-irrelevant sound automatically accesses the phonological loop, so that, English readings may be more distracted by unrelated sounds than Chinese readings.

Actually, there were two studies that also compared the effects of EL-BGM and CL-BGM on Chinese reading time in Chinese participants (Lu & Wang, 2019; Ren & Xu, 2019). Likewise, Lu and Wang (2019) found an interference effect of EL-BGM with insignificant but longer reaction time. On the contrary, Ren and Xu (2019) found a negative role for CL-BGM and a positive role for EL-BGM. Despite the fact that these two studies also employed within-subject designs and used different language versions of the same music excerpt, inconsistent findings were reached compared with the present study. The inconsistency may arise from participants’ English language proficiency. Ren and Xu (2019) used the scores of an English test as a screening criterion but did not analyze it as a covariate. Moreover, all participants of this study studied in a university where the medium of instruction is English for at least 2.5 years, so that, they may be more sensitive to English lyrics in the background music.

Theoretical considerations

The findings of this study did not support the interference-by-process hypothesis. Based on the hypothesis, because participants can understand Chinese and English, semantically processing the lyrics in these two languages should compete for the semantic processing that reading requires, so that, the performance should be impaired by background music in both Chinese and English. However, the results showed that performance was negatively affected by EL-BGM, but not CL-BGM. To assume the interference-by-process hypothesis is true and infer from it, the processing of CL-BGM may not or less depend on the semantic information. The semantic representation of lyrics in CL-BGM may undergo automatic processing. On the other hand, Martin et al. (1988) found that the phonological features of the to-be-ignored sound are not enough to distract comprehension. Therefore, the processing of CL-BGM may rely more on the phonological information, which consumes fewer cognitive resources that reading needs.

Meinhardt-Injac et al. (2015) claimed that such interference-by-process is a bottom-up mechanism because people encode the concurrent task and unrelated sound involuntarily. Hughes et al. (2013) suggested that the ISE results from not only bottom-up factors but also a top-down cognitive control, attentional capture. Salient irrelevant music takes people’s attention away from the primary task, modulating by task difficulty, foreknowledge, and working memory capacity. To extend this account, our data suggest that language proficiency also plays a role. English learners who were exposed to EL-BGM may allocate more attention to analyzing or translating the meaning of the English lyrics compared with CL-BGM, which may cause the distraction. This hypothesis may also explain another finding—the positive relationship between ELA and reading efficiency under the EL-BGM condition. With higher listening proficiency, less attention was occupied by processing the English lyrics’ meanings, so that, participants can pay more attention to comprehending the article.

Taking the above together and in light of the duplex-mechanism account (Hughes et al., 2013), the interference to cognitive tasks by shared processing irrelevant information (i.e., interference-by-process effect) can be reduced by top-down properties (e.g., language proficiency). In other words, the shared semantic processing of reading comprehension and intelligible lyrical music can cause ISE, but it depends on the proficiency of understanding the lyrics. High language proficiency of the task-unrelated lyrical music can leave the reading comprehension performance unaffected. Nonetheless, we should be aware that this assumption is based on an ideographic language (i.e., Chinese) reading. In addition, most research has shown a negative effect of CL-BGM on Chinese readings (Li & Li, 2006; Ren & Xu, 2019; Sun et al., 2011), which goes against this assumption. Thus, it needs further exploration and verifications.

Limitations and future directions

Several limitations are worth noting. First, the number of male and female participants was not balanced due to convenience sampling. Second, participants were recruited from the same university, so that, the sample was not representative. Third, the criteria of ELA were based on one listening test (with 10 questions) from the IELTS. It may not be reliable enough to demonstrate participants’ English listening proficiency. Further study can utilize two or more tests to ensure reliability.

Further study can ask Chinese English learners to complete a standardized English listening test, and divide them into three groups (i.e., low, medium, and high ELA) according to the score. As in the current study, each participant will exposed to three conditions (i.e., silence, CL-BGM, and EL-BGM) while reading Chinese articles. The purpose of this between–within-subject design is to examine whether there is an interaction between foreign language proficiency and the language of background music with lyrics on Chinese reading comprehension.

Implications

These findings have provided deeper insight into theories of ISE. In consideration of irrelevant intelligible music with lyrics, the effect may be first affected by top-down factors (e.g., language proficiency), then influenced by the similarity of processing between the task-related and task-unrelated information (interference-by-process theory). Our results also have practical implications for students and teachers. It is better not to listen to music with lyrics in English when reading a Chinese article if individuals can understand English but are not proficient in English listening.

Conclusion

Researchers have studied the impact of irrelevant sound on reading for eight decades. With the development of globalization, people tend to understand one or more languages other than their mother tongue. The present study opens the door to research about foreign language learners’ reading performance under exposure to background music with lyrics in two different intelligible languages, and their language proficiency is also taken into consideration. There were two major findings: (1) EL-BGM interfered with Chinese English learners’ Chinese reading comprehension performance, while CL-BGM did not and (2) with higher ELAs, English learners’ reading performance was less distracted under the EL-BGM condition.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded in part by the start-up research fund (R72021204) through BNU-HKBU United International College.

ORCID iD

Yi-Lung Kuo

Notes

References

AbyssMedia. (2014). Wave Editor (Version 3.3.3.0) [Computer software]. https://www.wave-editor.com/

AbyssMedia. (2016). AudioRetoucher (Version 3.9.1.0) [Computer software]. https://www.abyssmedia.com/

Anderson

Fuller

(2010). Effect of music on reading comprehension of junior high school students. School Psychology Quarterly, 25(3), 178–187. https://doi.org/10.1037/a0021213

Avila

Furnham

McClelland

(2011). The influence of distracting familiar vocal music on cognitive performance of introverts and extraverts. Psychology of Music, 40(1), 84–93. https://doi.org/10.1177/0305735611422672

Cambridge ESOL. (2009). Cambridge English IELTS 7 with answers: Examination papers from University of Cambridge ESOL Examinations. Cambridge University Press.

Chew

A. S. Q.

Y. T.

Chua

S. W.

Gan

S. K. E.

(2016). The effects of familiarity and language of background music on working memory and language tasks in Singapore. Psychology of Music, 44(6), 1431–1438. https://doi.org/10.1177/0305735616636209

Faul

Erdfelder

Buchner

Lang

A.-G.

(2009). Statistical power analyses using G*power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160. https://doi.org/10.3758/BRM.41.4.1149

Hallahan

D. P.

Kauflfman

J. M.

Pullen

P. C.

(2010). Exceptional learners: An introduction to special education ( Xiao

, Trans., 11th ed.). China Remin University Press.

Hughes

R. W.

Hurlstone

M. J.

Marsh

J. E.

Vachon

Jones

D. M.

(2013). Cognitive control of auditory distraction: Impact of task difficulty, foreknowledge, and working memory capacity supports duplex-mechanism account. Journal of Experimental Psychology: Human Perception and Performance, 39, 539–553. https://doi.org/10.1037/a0029064

10.

Hyönä

Ekholm

(2016). Background speech effects on sentence processing during reading: An eye movement study. PLOS ONE, 11(3), Article e0152133. https://doi.org/10.1371/journal.pone.0152133

11.

Kallinen

(2002). Reading news from a pocket computer in a distracting environment: Effects of the tempo of background music. Computers in Human Behavior, 18, 537–551. https://doi.org/10.1016/S0747-5632(02)00005-5

12.

N. N.

H. Y.

(2006). The effects of background music on the high school students’ reading comprehension. Studies of Psychology and Behavior, 4(2), 149–153. https://doi.org/10.4236/AP.2012.24032

13.

H. S.

Wang

(2019, October 18–20). The influence of Chinese and English background pop music to the Chinese and English reading comprehension in Chinese undergraduates [Paper presentation]. 22nd National Academic Congress of Psychology, Zhejiang, China.

14.

Martin

R. C.

Wogalter

M. S.

Forlano

J. G.

(1988). Reading comprehension in the presence of unattended speech and music. Journal of Memory and Language, 27(4), 382–398. https://doi.org/10.1016/0749-596X(88)90063-0

15.

Meinhardt-Injac

Schlittmeier

Klatte

Otto

Persike

Imhof

(2015). Auditory distraction by meaningless irrelevant speech: A developmental study. Applied Cognitive Psychology, 29, 217–225. https://doi.org/10.1002/acp.3098

16.

Perham

Currie

(2014). Does listening to preferred music improve reading comprehension performance? Applied Cognitive Psychology, 28, 279–284. https://doi.org/10.1002/acp.2994

17.

Ren

Y. N.

W. X.

(2019). Effect of Chinese and English background music on efficiency on Chinese and English reading comprehension. Advances in Psychology, 9(6), 978–984. https://doi.org/10.12677/AP.2019.96120

18.

Schneider

B. A.

Avivi-Reich

Mozuraitis

(2015). A cautionary note on the use of the analysis of covariance (ANCOVA) in classification designs with and without within-subject factors. Methods, 6, Article 474. https://doi.org/10.3389/fpsyg.2015.00474

19.

Schneider

Eschman

Zuccolotto

(2001). E-prime User’s Guide. Psychology Software Tools.

20.

Sun

N. N.

Zhou

T. H.

(2011). Popular music on students’ reading comprehension. China Journal of Health Psychology, 19(8), 978–980. https://doi.org/10.1017/CBO9781107415324.004

21.

Vasilev

M. R.

Kirkby

J. A.

Angele

(2018). Auditory distraction during reading: A Bayesian meta-analysis of a continuing controversy. Perspectives on Psychological Science, 13(5), 567–597. https://doi.org/10.1177/1745691617747398