Influential Factors for Medical Students’ Classroom Concentration—Evaluation With Speech Recognition and Face Recognition Technology

Influential Factors for Medical Students’ Classroom Concentration—Evaluation With Speech Recognition and Face Recognition Technology | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Influential Factors for Medical Students’ Classroom Concentration—Evaluation With Speech Recognition and Face Recognition Technology Xiaohan Chai, Jingwen Yang, Yunsong Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4600797/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 31 Oct, 2024 Read the published version in BMC Medical Education → Version 1 posted 4 You are reading this latest preprint version Abstract Statement of the Problem. The classroom concentration of medical students is an important factor in promoting their mastery of knowledge. Multiple teaching characteristics, such as speaking speed, voice volume, and question use, are confirmed to be influential factors. Purpose. This research aims to analyze how teachers’ linguistic characteristics affect medical students’ classroom concentration based on a speech recognition toolkit and face recognition technology. Materials and Methods : A speech recognition toolkit, WeNet, is used to recognize sentences during lectures in this study. Face recognition technology (FRT) is used to detect students’ concentration in class. The study involved 80 undergraduate students majoring in stomatology. The classroom videos of 5 class hours in the dental anatomy course were collected in October 2022. Pearson correlation, Spearman correlation and multiple linear regression analyses were used to analyze the impact of time and teachers’ linguistic characteristics on students’ concentration. Results. As a result of regression analysis, the explanatory power of the effect of the linguistic characteristics was 7.09% (F = 83.82, P < 0.001), with time, volume and question being significant influencing factors ( P < 0.01). The local polynomial smooth of the scatter between the concentration degree and the use of questions with time appears to fluctuate cyclically and suggests a potential inverse relationship between the use of questions and the concentration degree. Conclusions. The results of this study support the significant positive influence of volume and questioning technique, the negative influence of time, and the insignificant influence of speaking speed and the interval between sentences on students’ concentration. This study also suggested that teachers may adjust their questioning frequency based on their observation of students’ concentration. Figures Figure 1 CLINICAL IMPLICATIONS This study suggests that teachers can gain students’ attention through the way they speak during lectures based on evidence provided by artificial intelligence. Using questioning techniques and raising volume have a positive influence on students’ concentration. INTRODUCTION It is widely acknowledged that students’ classroom concentration toward lectures is crucial for achieving better mastery of knowledge. In recent years, an increasing number of factors have served as distractions to students due to the highly digitalized society 1 . Variations in teaching styles among teachers also play an important role in influencing students’ concentration 2 , 3 . Although pedagogies such as “problem-based learning (PBL)” and “generated question learning models (GQLM)” have been developed to improve students’ concentration, giving lectures still plays a predominant role in teaching forms 4 . Even under the same subject or based on similar pedagogies, there are still some teachers who can be more engaged with students in class, while others may not perform as well. By assessing students’ degree of concentration in response to different teaching characteristics, educators can gain valuable insights into the efficacy of various instructional approaches in promoting students’ concentration. Evaluating students’ classroom concentration has always been delayed because real-time evaluation can hardly be performed by methods such as self-reporting or testing, while students must “focus” on the lecture instead of judging their own concentrating state. As Bradbury mentioned, the data supporting the results of how students’ attention changes during a lecture were not satisfactory, as comments made by an individual human observer might not be as accurate as anticipated because of a lack of objectiveness 5 . In Bunce’s research, students reported lapses in attention after their occurrence using a clicker 6 . However, what the teacher did when lapses occurred was not considered. It is necessary to assess students’ reactions to the content given by teachers during lectures to determine more specific characteristics of teachers, which may reveal detailed techniques to help teachers give more engaging lectures. With the help of face recognition technology (FRT), students’ behavior and facial expression can be monitored objectively 7 , 8 . In 2023, artificial intelligence face recognition technology was used to evaluate medical students’ classroom concentration in real time 9 , 10 . Similar studies focused on both topic and teaching characteristics were performed in 2020, indicating that FRT is a reasonable and reliable tool for detecting student concentration 11 . However, the research presented in this article is insufficient to address our inquiries. Although this study explored the impact of lecture themes through the extraction of keywords via speech recognition, it did not mention the use of pedagogies. Furthermore, this study focused on heuristic classrooms in the field of physics, which demand more critical thinking from students, while memorizing classrooms in the field of medicine requires more understanding and memory from students. Thus, applying similar methods such as FRT and speech recognition technology in medical students’ class concentration remains necessary. In our study, we transform the lecture recordings into sentences to analyze linguistic characteristics. In 2022, an end-to-end speech recognition toolkit called WeNet was developed. WeNet is a productive toolkit for automatic speech recognition that combines the techniques of an n-gram-based language model, a unified contextual biasing framework and a unified IO system 12 . It is thought to be a productive toolkit with a lower error rate and to support large-scale datasets. WeNet is utilized in this study to accomplish the speech recognition process. With the help of WeNet, teachers’ linguistic characteristics can be further discussed. Many believe that the questioning technique is effective in promoting students’ concentration in class. A research article about questioning technique in medical education suggested that over half of faculties perceive questioning to be positive for students’ participation in class 13 . However, there is little quantitative evidence proving that the degree to which teachers use questions could significantly enhance students’ instance concentration. How fast the teacher speaks and how long the teacher pauses during a speech may influence students’ concentration. In 2023, Merhavy studied the influence of lecture playback speed on medical students’ concentration and memory using posttests and indicated that students’ learning ability might not be influenced by faster lecture playback speed 14 . Lenz’s study on teaching strategies suggested that using reflective pauses is beneficial for engaging students in class 13 . Thus, the influence of speaking speed and the interval between sentences are considered in this article. In Yang’s study, among the extracted audio features, audio volume is a factor that is more related to students’ concentration degree than to their speaking speed 11 . However, little evidence supports a similar conclusion, and further studies to determine the exact relationship between volume and concentration are needed. Another factor worth considering is the time of the lecture. Bradbury questioned the attention span of students during class, suggesting that more statistical evidence is needed to explain this topic 5 . Thus, this article takes time into consideration and analyses the relationship between the time after the beginning of a lecture and the change in students’ degree of concentration. In conclusion, with the help of WeNet and FRT, this study will discuss the impact of the pedagogy of using questioning techniques, the speaking speed of the teacher, the volume of the teacher’s voice during the lecture, and the time after the lecture begins on students’ concentration. MATERIALS AND METHODS Participants In 2022, 80 undergraduate students majoring in stomatology were selected as the research participants. The classroom videos of 5 class hours in the dental anatomy course were collected, and the videos lasted approximately 45 minutes each. All students and teachers in the course provided consent for this research (Approval number: PKUSSIRB-202274063). This study focuses on the relationship between students’ classroom concentration and 5 potentially related factors, including time after the lecture begins, the time interval between sentences, speaking speed, volume, and whether the teacher uses question forms to speak. Data collection In this study, the FRT, which was proven to be reliable in a previous study, was used to detect students’ facial states and analyze their degree of concentration 9 . A camera (A7S3, Sony, Japan) placed on the lectern is used to record students' classroom behaviors, which are set to high resolution with a low frame rate (4k, 30fps, 10bit). Each class session lasted for approximately 45 minutes. The images were exported in ".mp4" format in chronological order. After converting the ".mp4" files into images using the open-source video conversion software FFmpeg, one frame is extracted per second, resulting in a total of 21,600 images as the dataset for facial detection. The Face + + facial detection application programming interface (API) is employed to identify facial key points in the images. Study parameters The aim of this study was to focus on teachers’ linguistic characteristics, so sentences were used as the study unit. The characteristics of the sentences and how students’ concentrations changed after the sentences were analyzed. First, the verbal meaning of the sound in the videos was recognized by a speech recognition project called WeNet, an end-to-end speech recognition toolkit 12 . For the following parameters, each sentence was measured separately. 1. Students’ concentration degree The measurement of students’ concentration is based on FRT 9 . The number of faces judged as “concentrating” by FRT in each image is denoted as SF. The average SF for all the images in each sentence is noted as the “concentration degree (CD)”. 2. Time after the lecture began The time after the lecture begins is based on the exact start time of the sentence in the video. The unit is seconds. It is expressed as “time (T)” in this study. 3. Time interval between sentences This is defined as the time interval between the beginning of a sentence and the end of its former sentence. The unit of the time interval is seconds. It is expressed as “interval (I)” in this study. 4. Speaking speed Speaking speed represents how fast the teacher speaks. The unit of speaking speed is words (Chinese characters) per second. It is expressed as “speaking_speed (SS)”. 5. Volume In this article, we utilized the AudioSegment class from Python’s pydub package to read audio data and calculate its volume. By leveraging the start and end times of each sentence provided by WeNet, we extracted the audio data accordingly. The AudioSegment class comes with a built-in function for computing the volume of a segment of audio in decibels relative to full scale (dBFS). After reading the audio, the volume of the segment can be directly accessed as a member variable. This parameter is expressed as “volume (V)”. The volume of a sentence subtracting the volume of the former sentence is noted as the change in volume, which is considered to be another important key characteristic, expressed as “Vc”, with the unit of dBFS as well. 6. Using questioning expression or not Whether the teacher was using a questioning expression was also recorded. If WeNet decides that the sentence is interrogative, it will be marked as “1”; otherwise, it will be marked as “0”. The parameter is expressed as “question(Q)”. Statistical analysis The collected data were analyzed using the Stata/MP 17.0 program. In this study, Pearson correlation was used for continuous variables, including CD, T, I, SS, V and Vc, and Spearman correlation was used for nominal variables, including Q. Linear multiple regression was used to analyze the relationship between students’ concentration of each sentence, CD, and teachers’ linguistic characteristics. RESULTS The number of sentences recognized by WeNet and further taken into account was 3,308. In our analysis, the variable I did not meet the assumptions of normality. The skewness and kurtosis tests revealed significant differences from normality ( P < 0.05). Following the transformation of the square root of the variable, the distribution of I exhibited improved normality, as assessed by visual inspection and statistical tests. The descriptive statistical results are shown in Table 1 . The Pearson correlation coefficient was computed to examine the relationships between variables. Correlation analysis revealed a statistically significant negative correlation between CD and T stage (r = -0.2036, P < 0.001) and a statistically significant positive correlation between CD and V stage (r = 0.1717, P < 0.001). The correlation analysis also revealed a statistically significant positive correlation between Vc and V (r = 0.4779, P < 0.001). Spearman correlation coefficients were calculated to assess the relationships between questions and other variables, with Bonferroni correction for multiple comparisons. The analysis revealed a statistically significant positive correlation between questions and CD (Spearman's ρ = 0.1076, P < 0.001) and a marginally significant negative correlation between questions and T (Spearman's ρ = 0.0512, P = 0.0677). Table 1 Descriptive statistical results. Variable Obs Mean Std. dev. Min Max 1 CD 3308 8.660 3.022 1.970 22.817 2 T 3308 1609.919 940.757 0.000 3494.800 3 SS 3308 4.941 1.297 0.378 14.286 4 V 3308 -21.624 3.254 -34.645 -15.353 5 Vc 3303 0.000 3.115 -14.823 16.920 6 I 3303 0.614 0.409 0.000 3.378 7 Q 3308 0.182 0.386 0.000 1.000 1. CD = concentration degree; 2. T = time; 3. SS = speaking speed; 4. V = volume; 5. Vc = the change in the volume between a sentence and its previous sentence; 6. I = interval time between sentences; 7. Q = whether the teacher used questioning expression or not. Table 2 Regression analysis of concentration degree. 1 CD Coefficient std. err. t P > t [95% conf. interval] 2 T − .000624 .0000557 -11.20 0.000 -0.000733 -0.000515 3 V 0.141 0.016 8.84 0.000 0.110 0.173 4 Q 0.434 0.135 3.22 0.001 0.170 0.698 _cons 12.641 0.365 34.66 0.000 11.926 13.356 1. CD = concentration degree; 2. T = time; 3. V = volume; 4. Q = whether the teacher used questioning expression or not. As a result of regression analysis, the explanatory power of the effect of the linguistic characteristics was 7.09% (F = 83.82, P < 0.001), with T, volume and question being significant influencing factors ( P < 0.01) (Table 2 ). We computed the variance inflation factor (VIF) values for each predictor variable in the regression model, and the VIFs were all less than 5 (with a mean VIF of 1.03), indicating that collinearity did not occur within the predictor variables. The linear multiple regression equation is as follows: CD = -0.000624 T + 0.141 V + 0.434 Q + 12.641 Due to the observed fluctuating trend in the scatter plot of CD-T in Fig. 1 , we generated a local polynomial smooth plot using an Epanechnikov kernel function with a polynomial order of 6. The local smooth polynomials of CD-T and Q-T are depicted in Fig. 1 (a) and (b). The overlapping graph is depicted in Fig. 1 (c), in which the value of Q is 25 times larger to observe the two curves in one axis jointly. The local polynomial smooth of the scatter between CD and the question with T appears to cyclically fluctuate akin to that of periodic waves. The cycle of a period is approximately 10–15 minutes, and the general trend of CD appears to decrease during the lecture. The local polynomial smooth curves of CD and question appear to exhibit opposite trends, suggesting a potential inverse relationship between them. DISCUSSION In this study on the classroom concentration of medical students, the trends of students’ attention over time and the influence of teachers’ linguistic characteristics on concentration were discussed. The aim was to understand the relationships between various factors and to explore strategies that could improve students' classroom attention, both individually and collectively. The design of this study relies on the innovative application of two key technologies. The first one is the speech recognition toolkit called WeNet 12 . It is a great tool for educators to analyze large-scale datasets such as lecture recordings effectively. The other key technology is FRT, which was introduced in 2023 9 . This study provides further evidence supporting the use of FRT in analyzing students’ concentration. This study reveals several significant influential factors of teachers’ linguistic characteristics for students’ classroom attention through the analysis of sentences, including time, volume, and questioning techniques. Speaking speed and the interval between sentences and students’ concentration are considered to be nonsignificant influential factors. We also discuss the possible explanations of the curves of the changes in the concentrations of the questions and concentrations with time, which cyclically fluctuate, similar to periodic waves. The research findings indicate that, akin to common knowledge, students' classroom concentration tends to decline as the class progresses 13 . Some studies suggest that students’ attention could hardly last for more than 30 minutes during a lecture 15 , 16 . However, in Bunce’s statistical study, students pay attention in a shorter cycle of approximately 4.5 minutes, and the attention lapse occurs again at a shorter and shorter cycle through the lecture segment 6 . In this study, the relationship between the concentration and time was evidenced by an overall negative correlation between the two variables. According to the results of the scatter plots and polynomial smooth plots, this negative correlation may not be linear. Based on this observation, we conducted a practical exploration and found literature supporting the reproducibility of similar conclusions 5 , 17 . Based on the visual results in this study, the cycle of the concentration degree fluctuation period was approximately 10–15 minutes, with a decreasing trend. Another correlation finding indicates that volume is also a positive factor influencing concentration. This observation is consistent with previous literature, further corroborating each other's claims 11 . The statistical insignificance of certain correlations also holds practical significance. Previous research on the impact of video playback speed on students’ attention and memory retention did not find significant relationships, suggesting that playback speeds of ×1.5-2.0 do not interfere with learning outcomes 14 , 18 . In this study, the lack of significance between concentration and speaking speed and time interval further supports this conclusion. In addition, the factors examined in this study included the change in volume from one sentence to its previous sentence. It is commonly assumed that a lecture without variation in volume may be monotonous, while fluctuations in volume might be more engaging. However, this study did not find significant effects of volume change on the concentration. According to the FRT and WeNet results, the impact of the change in volume on the concentration degree is nonsignificant in our study, while the impact of volume is significantly positive. The results of the local polynomial smooth plots suggest that teachers may engage in unconscious subjective observations of students’ degree of concentration, relying on real-time feedback in their own mind as a potential basis for adjusting their questioning frequency in an attempt to motivate students to increase their level of attentiveness. When students are more focused, teachers may reduce their questioning frequency by employing fewer questioning strategies to advance classroom progress. As presented in the overlapping curves, if there is a causal relationship between the two variables, the response time of this effect should be relatively rapid. This study also has certain limitations. First, according to the results of multiple linear regression, with only 7.09% of the R-squared value indicating that although factors such as time, volume, and the use of questioning have a significant impact on students’ concentration, they do not have a decisive effect. This is understandable because factors such as the type and subject of the course, the use of electronic devices, and the application of teaching methods also have relatively certain influences on classroom concentration1. Due to the basic nature of the speech recognition research method used in this paper, these factors were not fully incorporated into the analysis. Our previous study reported a recall rate of 81.4%, which might have led to the introduction of errors in this study 9 . Therefore, although the FRT was considered reliable in our study, there is still potential for improving its accuracy. In the future, discussions on other factors affecting classroom attention should continue to expand. This study only conducted a preliminary analysis of the linguistic information of teachers during lectures, with relatively basic factors included. However, in today's context of promising developments in semantic recognition artificial intelligence, more information related to content meaning rather than simple speech features can also be incorporated into analysis 11 . This may be of significant reference value for teachers in class preparation and classroom strategy design. CONCLUSIONS The results of this study support the significant positive influence of volume and questioning technique, the negative influence of time, and the insignificant influence of speaking speed and the interval between sentences on students’ concentration. This study also suggested that teachers may adjust their questioning frequency based on their observation of students’ concentration. Declarations Ethics approval and consent to participate Approval number: PKUSSIRB-202274063 Consent for publication Not applicable Availability of data and material Not applicable Competing interests The authors declare that they have no competing interests. Funding This study was financially supported by grants from Teaching Reform Funding Project of Peking University School and Hospital of Stomatology (2022-PT-06), National Natural Science Foundation of China (No. 82201022) and Key Health Projects of Science and Technology Development of Lanzhou (No.2021002). Authors' contributions Xiaohan Chai: Data curation; Formal analysis; Writing - original draft Jingwen Yang: Funding acquisition; Investigation; Methodology; Project administration; Resources; Writing - review & editing Yunsong Liu : Project administration; Resources; Supervision; Validation; Writing - review & editing. Acknowledgements The authors thank Peking University School and Hospital of Stomatology & National Center for Stomatology & National Clinical Research Center for Oral Diseases & National Engineering Research Center of Oral Biomaterials and Digital Medical Devices& Beijing Key Laboratory of Digital Stomatology & NHC Key Laboratory of Digital Stomatology & NMPA Key Laboratory for Dental Materials for their assistance in preparing this manuscript. References Attia NA, Baig L, Marzouk YI, Khan A. The potential effect of technology and distractions on undergraduate students' concentration. Pak J Med Sci. 2017 Jul-Aug;33(4):860–5. https://doi.org/10.12669/pjms.334.12560 . Hake RR. Interactive-engagement versus traditional methods: A six-thousand-student survey of mechanics test data for introductory physics courses. Am J Phys. 1998;66(1):64–74. http://dx.doi.org/10.1119/1.18809 . Berman AC. Good teaching is good teaching: A narrative review for effective medical educators. Anat Sci Educ. 2015;8(4):386–94. https://doi.org/10.1002/ase.1535 . Miller CJ, McNear J, Metz MJ. A comparison of traditional and engaging lecture methods in a large, professional-level course. Adv Physiol Educ. 2013;37(4):347–55. https://doi.org/10.1152/advan.00050.2013 . Bradbury N. Do Students Really Have an Inability to Concentrate During Lectures? Acad Med. 2017;92(4):428–428. https://doi.org/10.1097/acm.0000000000001584 . Bunce DM, Flens EA, Neiles KY. How Long Can Students Pay Attention in Class? A Study of Student Attention Decline Using Clickers. J Chem Educ. 2010;87(12):1438–43. https://doi.org/10.1021/ed100409p . Whitehill J, Serpell Z, Lin YC, Foster A, Movellan JR. The Faces of Engagement: Automatic Recognition of Student Engagementfrom Facial Expressions. IEEE Trans Affect Comput. 2014;5(1):86–98. https://doi.org/10.1109/TAFFC.2014.2316163 . Zhang X, Zhang X, Jasni Bin Dolah TsD. Intelligent Classroom Teaching Assessment System Based on Deep Learning Model Face Recognition Technology. Sci Program. 2022;2022:1–10. https://doi.org/10.1155/2022%2F1851409 . Yang J, Yang Z, Li J, Liu Y. Study of artificial intelligence face recognition technology for real ̄time assessment of medical students′ classroom concentration. Chin J Med Educ. 2023;43(1):31–4. https://doi.org/10.3760/cma.j.cn115259-20220516-00628 . Tang J, Zhou X, Zheng J. Design of Intelligent classroom facial recognition based on Deep Learning. J Phys: Conf Ser. 2019;1168:022043. https://doi.org/10.1088/1742-6596/1168/2/022043 . Yang B, Yao Z, Lu H, Zhou Y, Xu J. In-classroom learning analytics based on student behavior, topic and teaching characteristic mining. Pattern Recognit Lett. 2020;129:224–31. https://doi.org/10.1016/j.patrec.2019.11.023 . Zhang B, Wu D, Peng Z, Song X, Yao Z, Lv H et al. WeNet 2.0: More Productive End-to-End Speech Recognition Toolkit [Internet]. arXiv; 2022 [cited 2024 Mar 22]. https://doi.org/10.48550/arXiv.2203.15455 . Lenz PH, McCallister JW, Luks AM, Le TT, Fessler HE. Practical Strategies for Effective Lectures. Am Thorac Soc. 2015;12(4):561–6. https://doi.org/10.1513/annalsats.201501-024ar . Merhavy ZI, Bassett L, Melchiorre M, Hall MPM. The impact of lecture playback speeds on concentration and memory. BMC Med Educ. 2023;23(1):515. https://doi.org/10.1186/s12909-023-04491-y . Hartley J, Davies IK. Note-taking: A critical review. Program Learn Educational Technol. 1978;15(3):207–24. https://doi.org/10.1080/0033039780150305 . Stuart J, Rutherford RJD. Medical student concentration during lectures. Lancet. 1978;312(8088):514–6. https://doi.org/10.1016/s0140-6736(78)92233-x . Bradbury NA. Attention span during lectures: 8 seconds, 10 minutes, or more? Adv Physiol Educ. 2016;40(4):509–13. https://doi.org/10.1152/advan.00109.2016 . Kıyak YS, Budakoğlu Iİ, Masters K, Coşkun Ö. The effect of watching lecture videos at 2× speed on memory retention performance of medical students: An experimental study. Med Teach. 2023;45(8):913–7. https://doi.org/10.1080/0142159x.2023.2189537 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 31 Oct, 2024 Read the published version in BMC Medical Education → Version 1 posted Editorial decision: Revision requested 24 Jun, 2024 Editor assigned by journal 20 Jun, 2024 Submission checks completed at journal 20 Jun, 2024 First submitted to journal 18 Jun, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4600797","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":318090502,"identity":"21b37f32-6e43-4f79-b28a-4a40190fc271","order_by":0,"name":"Xiaohan Chai","email":"","orcid":"","institution":"Peking University Stomatological Hospital","correspondingAuthor":false,"prefix":"","firstName":"Xiaohan","middleName":"","lastName":"Chai","suffix":""},{"id":318090503,"identity":"0e9ead7e-979e-4255-97c0-ad14473b5d29","order_by":1,"name":"Jingwen Yang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA90lEQVRIiWNgGAWjYBACxmZk3gcbsFjjAaK1MM5IA1MNeLWgAGaeNAgDrxbmdt7Dr3kq7jAYHD97+LVNgk3i2vbDQFtqbKJxO4wvzZrnzDMGgzN5adY5CWnGZmcSgVqOpeU24NTCY2ac23aYweBADpDx47Cc2QGgFsaGwwS0/ANqOf/GzNgi4TCP2fmHBLUYP85tAGq5kWP8mCEBaMsNImxh/nPsMI/kjTdmjD0gv9wA2pKAxy+G/WeMP86oOSzHdz7H+MMPYIhtO5/+8MGHGhvcWhoY2CSANI/CAQgDAhJwKAcBeWDUfAAzGqCMUTAKRsEoGAXoAACNtGMfb642VgAAAABJRU5ErkJggg==","orcid":"","institution":"Peking University Stomatological Hospital","correspondingAuthor":true,"prefix":"","firstName":"Jingwen","middleName":"","lastName":"Yang","suffix":""},{"id":318090504,"identity":"ad9566af-d530-45e5-b4d8-63b1cb25244a","order_by":2,"name":"Yunsong Liu","email":"","orcid":"","institution":"Peking University Stomatological Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yunsong","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2024-06-18 14:47:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4600797/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4600797/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12909-024-06204-5","type":"published","date":"2024-10-31T16:05:16+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":59978964,"identity":"57bb57c6-efde-4836-9e3a-f61e9962ff4d","added_by":"auto","created_at":"2024-07-10 05:33:51","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":182066,"visible":true,"origin":"","legend":"\u003cp\u003eLocal polynomial smooth of \u003csup\u003e1\u003c/sup\u003eCD-\u003csup\u003e2\u003c/sup\u003eT and \u003csup\u003e3\u003c/sup\u003eQ-T.\u003c/p\u003e\n\u003cp\u003ea. Local polynomial smooth of CD-T.\u003c/p\u003e\n\u003cp\u003eb. Local polynomial smooth of Q-T.\u003c/p\u003e\n\u003cp\u003ec. Overlap local polynomial smooth of CD-T and Q-T.\u003c/p\u003e\n\u003cp\u003e1. CD = concentration degree; 2. Q = whether the teacher usedquestioning expression; 3. T = time.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4600797/v1/daff4374017e056f15a4682f.png"},{"id":68207120,"identity":"85ed0f56-ecbd-4cb4-bf0c-cbe72b74392b","added_by":"auto","created_at":"2024-11-04 16:35:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":630482,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4600797/v1/44ab1694-0f76-4b8d-964b-78743d0d26de.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eInfluential Factors for Medical Students’ Classroom Concentration—Evaluation With Speech Recognition and Face Recognition Technology\u003c/p\u003e","fulltext":[{"header":"CLINICAL IMPLICATIONS","content":"\u003cp\u003eThis study suggests that teachers can gain students\u0026rsquo; attention through the way they speak during lectures based on evidence provided by artificial intelligence. Using questioning techniques and raising volume have a positive influence on students\u0026rsquo; concentration.\u003c/p\u003e"},{"header":"INTRODUCTION","content":"\u003cp\u003eIt is widely acknowledged that students\u0026rsquo; classroom concentration toward lectures is crucial for achieving better mastery of knowledge. In recent years, an increasing number of factors have served as distractions to students due to the highly digitalized society\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Variations in teaching styles among teachers also play an important role in influencing students\u0026rsquo; concentration\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eAlthough pedagogies such as \u0026ldquo;problem-based learning (PBL)\u0026rdquo; and \u0026ldquo;generated question learning models (GQLM)\u0026rdquo; have been developed to improve students\u0026rsquo; concentration, giving lectures still plays a predominant role in teaching forms\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Even under the same subject or based on similar pedagogies, there are still some teachers who can be more engaged with students in class, while others may not perform as well. By assessing students\u0026rsquo; degree of concentration in response to different teaching characteristics, educators can gain valuable insights into the efficacy of various instructional approaches in promoting students\u0026rsquo; concentration.\u003c/p\u003e \u003cp\u003eEvaluating students\u0026rsquo; classroom concentration has always been delayed because real-time evaluation can hardly be performed by methods such as self-reporting or testing, while students must \u0026ldquo;focus\u0026rdquo; on the lecture instead of judging their own concentrating state. As Bradbury mentioned, the data supporting the results of how students\u0026rsquo; attention changes during a lecture were not satisfactory, as comments made by an individual human observer might not be as accurate as anticipated because of a lack of objectiveness\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. In Bunce\u0026rsquo;s research, students reported lapses in attention after their occurrence using a clicker\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. However, what the teacher did when lapses occurred was not considered. It is necessary to assess students\u0026rsquo; reactions to the content given by teachers during lectures to determine more specific characteristics of teachers, which may reveal detailed techniques to help teachers give more engaging lectures.\u003c/p\u003e \u003cp\u003eWith the help of face recognition technology (FRT), students\u0026rsquo; behavior and facial expression can be monitored objectively\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. In 2023, artificial intelligence face recognition technology was used to evaluate medical students\u0026rsquo; classroom concentration in real time\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Similar studies focused on both topic and teaching characteristics were performed in 2020, indicating that FRT is a reasonable and reliable tool for detecting student concentration\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. However, the research presented in this article is insufficient to address our inquiries. Although this study explored the impact of lecture themes through the extraction of keywords via speech recognition, it did not mention the use of pedagogies. Furthermore, this study focused on heuristic classrooms in the field of physics, which demand more critical thinking from students, while memorizing classrooms in the field of medicine requires more understanding and memory from students. Thus, applying similar methods such as FRT and speech recognition technology in medical students\u0026rsquo; class concentration remains necessary.\u003c/p\u003e \u003cp\u003eIn our study, we transform the lecture recordings into sentences to analyze linguistic characteristics. In 2022, an end-to-end speech recognition toolkit called WeNet was developed. WeNet is a productive toolkit for automatic speech recognition that combines the techniques of an n-gram-based language model, a unified contextual biasing framework and a unified IO system\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. It is thought to be a productive toolkit with a lower error rate and to support large-scale datasets. WeNet is utilized in this study to accomplish the speech recognition process.\u003c/p\u003e \u003cp\u003eWith the help of WeNet, teachers\u0026rsquo; linguistic characteristics can be further discussed. Many believe that the questioning technique is effective in promoting students\u0026rsquo; concentration in class. A research article about questioning technique in medical education suggested that over half of faculties perceive questioning to be positive for students\u0026rsquo; participation in class\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. However, there is little quantitative evidence proving that the degree to which teachers use questions could significantly enhance students\u0026rsquo; instance concentration.\u003c/p\u003e \u003cp\u003eHow fast the teacher speaks and how long the teacher pauses during a speech may influence students\u0026rsquo; concentration. In 2023, Merhavy studied the influence of lecture playback speed on medical students\u0026rsquo; concentration and memory using posttests and indicated that students\u0026rsquo; learning ability might not be influenced by faster lecture playback speed\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. Lenz\u0026rsquo;s study on teaching strategies suggested that using reflective pauses is beneficial for engaging students in class\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Thus, the influence of speaking speed and the interval between sentences are considered in this article.\u003c/p\u003e \u003cp\u003eIn Yang\u0026rsquo;s study, among the extracted audio features, audio volume is a factor that is more related to students\u0026rsquo; concentration degree than to their speaking speed\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. However, little evidence supports a similar conclusion, and further studies to determine the exact relationship between volume and concentration are needed.\u003c/p\u003e \u003cp\u003eAnother factor worth considering is the time of the lecture. Bradbury questioned the attention span of students during class, suggesting that more statistical evidence is needed to explain this topic\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Thus, this article takes time into consideration and analyses the relationship between the time after the beginning of a lecture and the change in students\u0026rsquo; degree of concentration.\u003c/p\u003e \u003cp\u003eIn conclusion, with the help of WeNet and FRT, this study will discuss the impact of the pedagogy of using questioning techniques, the speaking speed of the teacher, the volume of the teacher\u0026rsquo;s voice during the lecture, and the time after the lecture begins on students\u0026rsquo; concentration.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eIn 2022, 80 undergraduate students majoring in stomatology were selected as the research participants. The classroom videos of 5 class hours in the dental anatomy course were collected, and the videos lasted approximately 45 minutes each. All students and teachers in the course provided consent for this research (Approval number: PKUSSIRB-202274063).\u003c/p\u003e \u003cp\u003eThis study focuses on the relationship between students\u0026rsquo; classroom concentration and 5 potentially related factors, including time after the lecture begins, the time interval between sentences, speaking speed, volume, and whether the teacher uses question forms to speak.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eData collection\u003c/h2\u003e \u003cp\u003eIn this study, the FRT, which was proven to be reliable in a previous study, was used to detect students\u0026rsquo; facial states and analyze their degree of concentration\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. A camera (A7S3, Sony, Japan) placed on the lectern is used to record students' classroom behaviors, which are set to high resolution with a low frame rate (4k, 30fps, 10bit). Each class session lasted for approximately 45 minutes. The images were exported in \".mp4\" format in chronological order.\u003c/p\u003e \u003cp\u003eAfter converting the \".mp4\" files into images using the open-source video conversion software FFmpeg, one frame is extracted per second, resulting in a total of 21,600 images as the dataset for facial detection. The Face\u0026thinsp;+\u0026thinsp;+\u0026thinsp;facial detection application programming interface (API) is employed to identify facial key points in the images.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eStudy parameters\u003c/h2\u003e \u003cp\u003eThe aim of this study was to focus on teachers\u0026rsquo; linguistic characteristics, so sentences were used as the study unit. The characteristics of the sentences and how students\u0026rsquo; concentrations changed after the sentences were analyzed. First, the verbal meaning of the sound in the videos was recognized by a speech recognition project called WeNet, an end-to-end speech recognition toolkit\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. For the following parameters, each sentence was measured separately.\u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e1. Students\u0026rsquo; concentration degree\u003c/h2\u003e \u003cp\u003eThe measurement of students\u0026rsquo; concentration is based on FRT\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. The number of faces judged as \u0026ldquo;concentrating\u0026rdquo; by FRT in each image is denoted as SF. The average SF for all the images in each sentence is noted as the \u0026ldquo;concentration degree (CD)\u0026rdquo;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e \u003ch2\u003e2. Time after the lecture began\u003c/h2\u003e \u003cp\u003eThe time after the lecture begins is based on the exact start time of the sentence in the video. The unit is seconds. It is expressed as \u0026ldquo;time (T)\u0026rdquo; in this study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e3. Time interval between sentences\u003c/h2\u003e \u003cp\u003eThis is defined as the time interval between the beginning of a sentence and the end of its former sentence. The unit of the time interval is seconds. It is expressed as \u0026ldquo;interval (I)\u0026rdquo; in this study.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e4. Speaking speed\u003c/h2\u003e \u003cp\u003eSpeaking speed represents how fast the teacher speaks. The unit of speaking speed is words (Chinese characters) per second. It is expressed as \u0026ldquo;speaking_speed (SS)\u0026rdquo;.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e5. Volume\u003c/h2\u003e \u003cp\u003eIn this article, we utilized the AudioSegment class from Python\u0026rsquo;s pydub package to read audio data and calculate its volume. By leveraging the start and end times of each sentence provided by WeNet, we extracted the audio data accordingly. The AudioSegment class comes with a built-in function for computing the volume of a segment of audio in decibels relative to full scale (dBFS). After reading the audio, the volume of the segment can be directly accessed as a member variable. This parameter is expressed as \u0026ldquo;volume (V)\u0026rdquo;.\u003c/p\u003e \u003cp\u003eThe volume of a sentence subtracting the volume of the former sentence is noted as the change in volume, which is considered to be another important key characteristic, expressed as \u0026ldquo;Vc\u0026rdquo;, with the unit of dBFS as well.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e6. Using questioning expression or not\u003c/h2\u003e \u003cp\u003eWhether the teacher was using a questioning expression was also recorded. If WeNet decides that the sentence is interrogative, it will be marked as \u0026ldquo;1\u0026rdquo;; otherwise, it will be marked as \u0026ldquo;0\u0026rdquo;. The parameter is expressed as \u0026ldquo;question(Q)\u0026rdquo;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eThe collected data were analyzed using the Stata/MP 17.0 program. In this study, Pearson correlation was used for continuous variables, including CD, T, I, SS, V and Vc, and Spearman correlation was used for nominal variables, including Q. Linear multiple regression was used to analyze the relationship between students\u0026rsquo; concentration of each sentence, CD, and teachers\u0026rsquo; linguistic characteristics.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003eThe number of sentences recognized by WeNet and further taken into account was 3,308. In our analysis, the variable I did not meet the assumptions of normality. The skewness and kurtosis tests revealed significant differences from normality (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Following the transformation of the square root of the variable, the distribution of I exhibited improved normality, as assessed by visual inspection and statistical tests. The descriptive statistical results are shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe Pearson correlation coefficient was computed to examine the relationships between variables. Correlation analysis revealed a statistically significant negative correlation between CD and T stage (r = -0.2036, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and a statistically significant positive correlation between CD and V stage (r\u0026thinsp;=\u0026thinsp;0.1717, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The correlation analysis also revealed a statistically significant positive correlation between Vc and V (r\u0026thinsp;=\u0026thinsp;0.4779, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eSpearman correlation coefficients were calculated to assess the relationships between questions and other variables, with Bonferroni correction for multiple comparisons. The analysis revealed a statistically significant positive correlation between questions and CD (Spearman's ρ\u0026thinsp;=\u0026thinsp;0.1076, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and a marginally significant negative correlation between questions and T (Spearman's ρ\u0026thinsp;=\u0026thinsp;0.0512, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0677).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDescriptive statistical results.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eObs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eStd. dev.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eMin\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMax\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eCD\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e8.660\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.970\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e22.817\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eT\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1609.919\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e940.757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3494.800\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eSS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.941\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.297\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.378\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e14.286\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eV\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e-21.624\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.254\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-34.645\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-15.353\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eVc\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3303\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.115\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e-14.823\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e16.920\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3303\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.614\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3.378\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eQ\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3308\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.182\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.386\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e1. CD\u0026thinsp;=\u0026thinsp;concentration degree; 2. T\u0026thinsp;=\u0026thinsp;time; 3. SS\u0026thinsp;=\u0026thinsp;speaking speed; 4. V\u0026thinsp;=\u0026thinsp;volume; 5. Vc\u0026thinsp;=\u0026thinsp;the change in the volume between a sentence and its previous sentence; 6. I\u0026thinsp;=\u0026thinsp;interval time between sentences; 7. Q\u0026thinsp;=\u0026thinsp;whether the teacher used questioning expression or not.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRegression analysis of concentration degree.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003eCD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoefficient\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003estd. err.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003et\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP\u0026thinsp;\u0026gt;\u0026thinsp;t\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003e[95% conf. interval]\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eT\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;\u0026thinsp;.000624\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e.0000557\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-11.20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e-0.000733\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e-0.000515\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eV\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.141\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.173\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003csup\u003e\u003cb\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/b\u003e\u003c/sup\u003e\u003cb\u003eQ\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.434\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.135\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.170\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.698\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e_cons\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.641\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.365\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e34.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e11.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e13.356\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e1. CD\u0026thinsp;=\u0026thinsp;concentration degree; 2. T\u0026thinsp;=\u0026thinsp;time; 3. V\u0026thinsp;=\u0026thinsp;volume; 4. Q\u0026thinsp;=\u0026thinsp;whether the teacher used questioning expression or not.\u003c/p\u003e \u003cp\u003eAs a result of regression analysis, the explanatory power of the effect of the linguistic characteristics was 7.09% (F\u0026thinsp;=\u0026thinsp;83.82, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with T, volume and question being significant influencing factors (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). We computed the variance inflation factor (VIF) values for each predictor variable in the regression model, and the VIFs were all less than 5 (with a mean VIF of 1.03), indicating that collinearity did not occur within the predictor variables. The linear multiple regression equation is as follows:\u003c/p\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003cp\u003eCD = -0.000624 T\u0026thinsp;+\u0026thinsp;0.141 V\u0026thinsp;+\u0026thinsp;0.434 Q\u0026thinsp;+\u0026thinsp;12.641\u003c/p\u003e \u003cp\u003eDue to the observed fluctuating trend in the scatter plot of CD-T in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, we generated a local polynomial smooth plot using an Epanechnikov kernel function with a polynomial order of 6. The local smooth polynomials of CD-T and Q-T are depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e(a) and (b). The overlapping graph is depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e(c), in which the value of Q is 25 times larger to observe the two curves in one axis jointly.\u003c/p\u003e \u003cp\u003eThe local polynomial smooth of the scatter between CD and the question with T appears to cyclically fluctuate akin to that of periodic waves. The cycle of a period is approximately 10\u0026ndash;15 minutes, and the general trend of CD appears to decrease during the lecture.\u003c/p\u003e \u003cp\u003eThe local polynomial smooth curves of CD and question appear to exhibit opposite trends, suggesting a potential inverse relationship between them.\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eIn this study on the classroom concentration of medical students, the trends of students\u0026rsquo; attention over time and the influence of teachers\u0026rsquo; linguistic characteristics on concentration were discussed. The aim was to understand the relationships between various factors and to explore strategies that could improve students' classroom attention, both individually and collectively.\u003c/p\u003e \u003cp\u003eThe design of this study relies on the innovative application of two key technologies. The first one is the speech recognition toolkit called WeNet\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. It is a great tool for educators to analyze large-scale datasets such as lecture recordings effectively. The other key technology is FRT, which was introduced in 2023\u003csup\u003e9\u003c/sup\u003e. This study provides further evidence supporting the use of FRT in analyzing students\u0026rsquo; concentration.\u003c/p\u003e \u003cp\u003eThis study reveals several significant influential factors of teachers\u0026rsquo; linguistic characteristics for students\u0026rsquo; classroom attention through the analysis of sentences, including time, volume, and questioning techniques. Speaking speed and the interval between sentences and students\u0026rsquo; concentration are considered to be nonsignificant influential factors. We also discuss the possible explanations of the curves of the changes in the concentrations of the questions and concentrations with time, which cyclically fluctuate, similar to periodic waves.\u003c/p\u003e \u003cp\u003eThe research findings indicate that, akin to common knowledge, students' classroom concentration tends to decline as the class progresses\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Some studies suggest that students\u0026rsquo; attention could hardly last for more than 30 minutes during a lecture\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. However, in Bunce\u0026rsquo;s statistical study, students pay attention in a shorter cycle of approximately 4.5 minutes, and the attention lapse occurs again at a shorter and shorter cycle through the lecture segment\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. In this study, the relationship between the concentration and time was evidenced by an overall negative correlation between the two variables. According to the results of the scatter plots and polynomial smooth plots, this negative correlation may not be linear. Based on this observation, we conducted a practical exploration and found literature supporting the reproducibility of similar conclusions\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Based on the visual results in this study, the cycle of the concentration degree fluctuation period was approximately 10\u0026ndash;15 minutes, with a decreasing trend.\u003c/p\u003e \u003cp\u003eAnother correlation finding indicates that volume is also a positive factor influencing concentration. This observation is consistent with previous literature, further corroborating each other's claims\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe statistical insignificance of certain correlations also holds practical significance. Previous research on the impact of video playback speed on students\u0026rsquo; attention and memory retention did not find significant relationships, suggesting that playback speeds of \u0026times;1.5-2.0 do not interfere with learning outcomes\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. In this study, the lack of significance between concentration and speaking speed and time interval further supports this conclusion.\u003c/p\u003e \u003cp\u003eIn addition, the factors examined in this study included the change in volume from one sentence to its previous sentence. It is commonly assumed that a lecture without variation in volume may be monotonous, while fluctuations in volume might be more engaging. However, this study did not find significant effects of volume change on the concentration. According to the FRT and WeNet results, the impact of the change in volume on the concentration degree is nonsignificant in our study, while the impact of volume is significantly positive.\u003c/p\u003e \u003cp\u003eThe results of the local polynomial smooth plots suggest that teachers may engage in unconscious subjective observations of students\u0026rsquo; degree of concentration, relying on real-time feedback in their own mind as a potential basis for adjusting their questioning frequency in an attempt to motivate students to increase their level of attentiveness. When students are more focused, teachers may reduce their questioning frequency by employing fewer questioning strategies to advance classroom progress. As presented in the overlapping curves, if there is a causal relationship between the two variables, the response time of this effect should be relatively rapid.\u003c/p\u003e \u003cp\u003eThis study also has certain limitations. First, according to the results of multiple linear regression, with only 7.09% of the R-squared value indicating that although factors such as time, volume, and the use of questioning have a significant impact on students\u0026rsquo; concentration, they do not have a decisive effect. This is understandable because factors such as the type and subject of the course, the use of electronic devices, and the application of teaching methods also have relatively certain influences on classroom concentration1. Due to the basic nature of the speech recognition research method used in this paper, these factors were not fully incorporated into the analysis. Our previous study reported a recall rate of 81.4%, which might have led to the introduction of errors in this study\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Therefore, although the FRT was considered reliable in our study, there is still potential for improving its accuracy.\u003c/p\u003e \u003cp\u003eIn the future, discussions on other factors affecting classroom attention should continue to expand. This study only conducted a preliminary analysis of the linguistic information of teachers during lectures, with relatively basic factors included. However, in today's context of promising developments in semantic recognition artificial intelligence, more information related to content meaning rather than simple speech features can also be incorporated into analysis\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. This may be of significant reference value for teachers in class preparation and classroom strategy design.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eThe results of this study support the significant positive influence of volume and questioning technique, the negative influence of time, and the insignificant influence of speaking speed and the interval between sentences on students\u0026rsquo; concentration. This study also suggested that teachers may adjust their questioning frequency based on their observation of students\u0026rsquo; concentration.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eApproval number: PKUSSIRB-202274063\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was financially supported by grants from Teaching Reform Funding Project of Peking University School and Hospital of Stomatology (2022-PT-06), National Natural Science Foundation of China (No. 82201022) and Key Health Projects of Science and Technology Development of Lanzhou (No.2021002).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXiaohan Chai: Data curation; Formal analysis; Writing - original draft\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eJingwen Yang: Funding acquisition; Investigation; Methodology; Project administration; Resources; Writing - review \u0026amp; editing\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYunsong Liu : Project administration; Resources; Supervision; Validation; Writing - review \u0026amp; editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank Peking University School and Hospital of Stomatology \u0026amp; National Center for Stomatology \u0026amp; National Clinical Research Center for Oral Diseases \u0026amp; \u0026nbsp;National Engineering Research Center of Oral Biomaterials and Digital Medical Devices\u0026amp; Beijing Key Laboratory of Digital Stomatology \u0026amp; NHC Key Laboratory of Digital Stomatology \u0026amp; NMPA Key Laboratory for Dental Materials for their assistance in preparing this manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAttia NA, Baig L, Marzouk YI, Khan A. The potential effect of technology and distractions on undergraduate students' concentration. Pak J Med Sci. 2017 Jul-Aug;33(4):860\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.12669/pjms.334.12560\u003c/span\u003e\u003cspan address=\"10.12669/pjms.334.12560\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHake RR. Interactive-engagement versus traditional methods: A six-thousand-student survey of mechanics test data for introductory physics courses. Am J Phys. 1998;66(1):64\u0026ndash;74. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1119/1.18809\u003c/span\u003e\u003cspan address=\"10.1119/1.18809\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBerman AC. Good teaching is good teaching: A narrative review for effective medical educators. Anat Sci Educ. 2015;8(4):386\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ase.1535\u003c/span\u003e\u003cspan address=\"10.1002/ase.1535\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiller CJ, McNear J, Metz MJ. A comparison of traditional and engaging lecture methods in a large, professional-level course. Adv Physiol Educ. 2013;37(4):347\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1152/advan.00050.2013\u003c/span\u003e\u003cspan address=\"10.1152/advan.00050.2013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBradbury N. Do Students Really Have an Inability to Concentrate During Lectures? Acad Med. 2017;92(4):428\u0026ndash;428. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1097/acm.0000000000001584\u003c/span\u003e\u003cspan address=\"10.1097/acm.0000000000001584\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBunce DM, Flens EA, Neiles KY. How Long Can Students Pay Attention in Class? A Study of Student Attention Decline Using Clickers. J Chem Educ. 2010;87(12):1438\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1021/ed100409p\u003c/span\u003e\u003cspan address=\"10.1021/ed100409p\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWhitehill J, Serpell Z, Lin YC, Foster A, Movellan JR. The Faces of Engagement: Automatic Recognition of Student Engagementfrom Facial Expressions. IEEE Trans Affect Comput. 2014;5(1):86\u0026ndash;98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TAFFC.2014.2316163\u003c/span\u003e\u003cspan address=\"10.1109/TAFFC.2014.2316163\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang X, Zhang X, Jasni Bin Dolah TsD. Intelligent Classroom Teaching Assessment System Based on Deep Learning Model Face Recognition Technology. Sci Program. 2022;2022:1\u0026ndash;10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1155/2022%2F1851409\u003c/span\u003e\u003cspan address=\"10.1155/2022%2F1851409\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang J, Yang Z, Li J, Liu Y. Study of artificial intelligence face recognition technology for real ̄time assessment of medical students\u0026prime; classroom concentration. Chin J Med Educ. 2023;43(1):31\u0026ndash;4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3760/cma.j.cn115259-20220516-00628\u003c/span\u003e\u003cspan address=\"10.3760/cma.j.cn115259-20220516-00628\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang J, Zhou X, Zheng J. Design of Intelligent classroom facial recognition based on Deep Learning. J Phys: Conf Ser. 2019;1168:022043. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1088/1742-6596/1168/2/022043\u003c/span\u003e\u003cspan address=\"10.1088/1742-6596/1168/2/022043\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang B, Yao Z, Lu H, Zhou Y, Xu J. In-classroom learning analytics based on student behavior, topic and teaching characteristic mining. Pattern Recognit Lett. 2020;129:224\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.patrec.2019.11.023\u003c/span\u003e\u003cspan address=\"10.1016/j.patrec.2019.11.023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang B, Wu D, Peng Z, Song X, Yao Z, Lv H et al. WeNet 2.0: More Productive End-to-End Speech Recognition Toolkit [Internet]. arXiv; 2022 [cited 2024 Mar 22]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2203.15455\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2203.15455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLenz PH, McCallister JW, Luks AM, Le TT, Fessler HE. Practical Strategies for Effective Lectures. Am Thorac Soc. 2015;12(4):561\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1513/annalsats.201501-024ar\u003c/span\u003e\u003cspan address=\"10.1513/annalsats.201501-024ar\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMerhavy ZI, Bassett L, Melchiorre M, Hall MPM. The impact of lecture playback speeds on concentration and memory. BMC Med Educ. 2023;23(1):515. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s12909-023-04491-y\u003c/span\u003e\u003cspan address=\"10.1186/s12909-023-04491-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHartley J, Davies IK. Note-taking: A critical review. Program Learn Educational Technol. 1978;15(3):207\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0033039780150305\u003c/span\u003e\u003cspan address=\"10.1080/0033039780150305\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStuart J, Rutherford RJD. Medical student concentration during lectures. Lancet. 1978;312(8088):514\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/s0140-6736(78)92233-x\u003c/span\u003e\u003cspan address=\"10.1016/s0140-6736(78)92233-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBradbury NA. Attention span during lectures: 8 seconds, 10 minutes, or more? Adv Physiol Educ. 2016;40(4):509\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1152/advan.00109.2016\u003c/span\u003e\u003cspan address=\"10.1152/advan.00109.2016\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKıyak YS, Budakoğlu Iİ, Masters K, Coşkun \u0026Ouml;. The effect of watching lecture videos at 2\u0026times; speed on memory retention performance of medical students: An experimental study. Med Teach. 2023;45(8):913\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/0142159x.2023.2189537\u003c/span\u003e\u003cspan address=\"10.1080/0142159x.2023.2189537\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-education","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meed","sideBox":"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meed/default.aspx","title":"BMC Medical Education","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4600797/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4600797/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eStatement of the Problem.\u003c/strong\u003e The classroom concentration of medical students is an important factor in promoting their mastery of knowledge. Multiple teaching characteristics, such as speaking speed, voice volume, and question use, are confirmed to be influential factors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePurpose.\u003c/strong\u003e This research aims to analyze how teachers’ linguistic characteristics affect medical students’ classroom concentration based on a speech recognition toolkit and face recognition technology.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMaterials and Methods\u003c/strong\u003e: A speech recognition toolkit, WeNet, is used to recognize sentences during lectures in this study. Face recognition technology (FRT) is used to detect students’ concentration in class. The study involved 80 undergraduate students majoring in stomatology. The classroom videos of 5 class hours in the dental anatomy course were collected in October 2022. Pearson correlation, Spearman correlation and multiple linear regression analyses were used to analyze the impact of time and teachers’ linguistic characteristics on students’ concentration.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults.\u003c/strong\u003e As a result of regression analysis, the explanatory power of the effect of the linguistic characteristics was 7.09% (F = 83.82, \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001), with time, volume and question being significant influencing factors (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.01). The local polynomial smooth of the scatter between the concentration degree and the use of questions with time appears to fluctuate cyclically and suggests a potential inverse relationship between the use of questions and the concentration degree.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions.\u003c/strong\u003e The results of this study support the significant positive influence of volume and questioning technique, the negative influence of time, and the insignificant influence of speaking speed and the interval between sentences on students’ concentration. This study also suggested that teachers may adjust their questioning frequency based on their observation of students’ concentration.\u003c/p\u003e","manuscriptTitle":"Influential Factors for Medical Students’ Classroom Concentration—Evaluation With Speech Recognition and Face Recognition Technology","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-10 05:33:46","doi":"10.21203/rs.3.rs-4600797/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-06-24T05:45:19+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-06-20T11:45:17+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-06-20T11:44:52+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Education","date":"2024-06-18T14:46:09+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-education","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"meed","sideBox":"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/meed/default.aspx","title":"BMC Medical Education","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f6bcaead-6532-49d1-ad96-26da8fa6af18","owner":[],"postedDate":"July 10th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-11-04T16:24:38+00:00","versionOfRecord":{"articleIdentity":"rs-4600797","link":"https://doi.org/10.1186/s12909-024-06204-5","journal":{"identity":"bmc-medical-education","isVorOnly":false,"title":"BMC Medical Education"},"publishedOn":"2024-10-31 16:05:16","publishedOnDateReadable":"October 31st, 2024"},"versionCreatedAt":"2024-07-10 05:33:46","video":"","vorDoi":"10.1186/s12909-024-06204-5","vorDoiUrl":"https://doi.org/10.1186/s12909-024-06204-5","workflowStages":[]},"version":"v1","identity":"rs-4600797","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4600797","identity":"rs-4600797","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}