Subjective Tests Research Articles

Mel-S3R: Combining Mel-spectrogram and self-supervised speech representation with VQ-VAE for any-to-any voice conversion

The self-supervised speech representation (S3R) has succeeded in many downstream tasks, such as speaker recognition and voice conversion thanks to its high-level information. Voice conversion (VC) is a task to convert the source speech into a target speaker’s voice. Though S3R features effectively encode content and speaker information, spectral features contain low-level acoustic information that is complementary to the S3R. As a result, solely relying on the S3R features for VC may not be optimal. In order to seek speech representation carrying both high-level learned information and low-level spectral details for VC, we proposed a three-level attention to combine Mel-spectrogram (Mel) and S3R, denoted as Mel-S3R. In particular, S3R features are high-level learned representations extracted by a pre-trained network with self-supervised learning. Whereas Mel is the spectral feature representing the acoustic information. Then the proposed Mel-S3R is used as the input of any-to-any VQ-VAE-based VC and the experiments are performed as a downstream task. Objective metrics and subjective listening tests have demonstrated that the proposed Mel-S3R speech representation facilitates the VC framework to achieve robust performance in terms of both speech quality and speaker similarity.

Audio–Visual Cross-Modal Correspondences of Perceived Urgency: Examination through a Speeded Discrimination Task

Abstract When presenting information in vehicle cockpits, it is essential to convey an appropriate urgency to the drivers. Perceived urgency has been investigated over the years for each modality, particularly audition and vision. However, the interaction between the modalities of perceived urgency has rarely been examined. To expand the insight into the design application of information presentation, we investigated the audio–visual interaction of perceived urgency using a priming task that involved speeded visual-target discrimination. A total of 60 auditory stimuli were created using a synthesizer. In addition, 13 color squares were used as visual stimuli. Three auditory stimuli with high, medium, and low perceived urgency, and two visual stimuli with high and low perceived urgency were selected based on a subjective evaluation test using a seven-point scale. A priming task was conducted to examine the cross-modal interaction of perceived urgency. Auditory stimuli were presented as prime stimuli, and the participants were asked to discriminate the visual target as quickly as possible. The results revealed that auditory stimuli with high and low perceived urgency facilitated responses to each visual stimulus with similar perceived urgency relative to each visual stimulus with different perceived urgency. The auditory stimulus with medium perceived urgency also facilitated responses to the visual stimulus with high perceived urgency relative to low. The present study shows that cross-modal correspondences can be observed when the stimuli are selected based on their subjective perceived urgency.

Feasibility of measuring fusional vergence amplitudes objectively.

Two tests to measure fusional vergence amplitudes objectively were developed and validated against the two conventional clinical tests. Forty-nine adults participated in the study. Participants' negative (BI, base in) and positive (BO, base out) fusional vergence amplitudes at near were measured objectively in an haploscopic set-up by recording eye movements with an EyeLink 1000 Plus (SR Research). Stimulus disparity changed in steps or smoothly mimicking a prim bar and a Risley prism, respectively. Break and recovery points were determined offline using a custom Matlab algorithm for the analysis of eye movements. Fusional vergence amplitudes were also measured with two clinical tests using a Risley prism and a prism bar. A better agreement between tests was found for the measurement of BI than for BO fusional vergence amplitudes. The means ± SD of the differences between the BI break and recovery points measured with the two objective tests were -1.74 ± 3.35 PD and -1.97 ± 2.60 PD, respectively, which were comparable to those obtained for the subjective tests. For the BO break and recovery points, although the means of the differences between the two objective tests were small, high variability between subjects was found (0.31 ± 6.44 PD and -2.84 ± 7.01 PD, respectively). This study showed the feasibility to measure fusional vergence amplitudes objectively and overcome limitations of the conventional subjective tests. However, these tests cannot be used interchangeably due to their poor agreement.

Olfactory dysfunction in patients with Wilson's Disease.

Many neurodegenerative disorders are associated with olfactory dysfunction (OD), but little is known about OD in Wilson's Disease (WD). We evaluated olfactory function in patients with WD. OD was examined in 68 patients with WD and 70 sex- and age-matched healthy controls using subjective testing with 'Sniffin Sticks'. Threshold discrimination identification (TDI) score and its three components (odour detection threshold, discrimination, and identification) were assessed. Compared to controls, patients with WD had a significantly weaker sense of smell in terms of TDI (p < 0.01), odour discrimination (p < 0.01), and identification (p < 0.01), but not in terms of odour detection threshold (p = 0.27). Patients with predominantly neurological symptoms were characterised by greater OD by TDI (p < 0.01), odour detection threshold (p = 0.01), and discrimination (p = 0.03). The presence of pathological lesions (p = 0.04) in brain magnetic resonance imaging and generalised brain atrophy (p = 0.02) predisposed to worse TDI. In the WD group, weak inverse correlations between age and TDI score (r = -0.27), odour detection threshold (r = -0.3), and discrimination (r = -0.3) were found. Male gender was a risk factor for abnormal TDI in both WD and controls (both p = 0.02). Patients with WD, particularly older individuals, more frequently had OD than healthy volunteers. Predominantly neurological symptoms, and the presence of typical brain MRI changes, predisposed patients with WD to smell disorders.

Active noise control with selective perceptual equalization to shape the residual sound

An active noise control (ANC) system usually contains a control filter to generate an anti-noise sound that destructively interferes with the primary noise. The filtered reference least mean squares (FxLMS) algorithm is the seminal adaptive ANC algorithm that updates the control filter coefficients iteratively. However, the FxLMS algorithm is primarily designed to minimize the power of the residual sound, without consideration to the auditory sensation. Thus, the residual sound can be perceived to be different from the primary noise of a broad frequency band. When an ANC system running the FxLMS algorithm was piloted in a substation’s control room, workers expressed discomfort with the residual sound, even though the noise level was reduced. It was mainly because the workers were annoyed by the residual sound which they identified as an anomaly operating state. In order to resolve this problem, this paper proposes a selective perceptual equalization FxLMS (SPE-FxLMS) algorithm to retain the operating state in the residual sound. The simulation, experiment and subjective test results validate the effectiveness of the SPE-FxLMS algorithm in reducing the noise level and shaping the residual sound to create a sense of auditory familiarity for the workers.

Effectiveness of buffered and non-buffered local anaesthetic in inferior alveolar nerve block: a randomised study

The objective of this study was to compare the effectiveness of buffered and non-buffered local anaesthetic (LA) following inferior alveolar nerve block (IANB). This study was conducted at Usmanu Danfodiyo University Teaching Hospital Sokoto from June 2020 to January 2021. Subjects were randomised into Groups A and B. Group A received 2 ml of freshly prepared 2% lignocaine with 1:100,000 adrenaline buffered with 0.18 ml of 8.4% sodium bicarbonate solution while those in Group B received non-buffered 2% lignocaine with 1:100,000 adrenaline LA solutions. The onset of action of the LA was assessed by subjective and objective methods, while pain at the injection site was assessed using a numerical rating scale. Data obtained were analysed using statistical package for social sciences (IBM SPSS) version 21. The mean (SD) ages for Groups A and B were 37.4 (14.9) and 40.1 (14.4) years, respectively. The mean (SD) onset times of the LA by subjective testing were 126 (31.7) and 201 (66.8) seconds for Groups A and B, respectively. Similarly, the mean (SD) onset times of the local anaesthesia by objective testing for Groups A and B were 186 (41.0) and 287 (85.0) seconds, respectively, and both were significant (p<0.001). Pain at the injection site assessed objectively and subjectively were also statistically different (p<0.001). The findings of this study suggest that buffered LA is more effective than non-buffered LA of the same composition when used for IANB particularly concerning a significantly faster rate of onset and less pain at the injection site.

Evidence of vestibular dysfunction in children with enlarged vestibular aqueduct

ObjectiveTo investigate the occurrence and characteristics of balance and vestibular deficits in pediatric patients with enlarged vestibular aqueduct (EVA). Materials and methodsRetrospective review of 53 children with EVA who underwent a comprehensive vestibular evaluation in our pediatric balance and vestibular program. Laboratory testing included videonystagmography (VNG), rotary chair, video head impulse testing (vHIT), vestibular evoked myogenic potential (VEMP), subjective visual vertical (SVV) and Sensory Organization Test (SOT) in posturography. ResultsThe mean age of these children, 31 girls and 22 boys, was 7.1 years (SD = 4.8). Among these 53 children, 16 had unilateral EVA (7 on the left side and 9 on the right side) and 37 had bilateral EVA, in which genetic testing confirmed 5 cases of Pendred syndrome. Abnormal testing results were found in 58% (11/19) on SOT, 67% (32/48) on rotary chair, 55% (48/88 of ears) on VEMP, 30% (8/27) on vHIT, 39% (7/18) on SVV, and 8% (4/53) on VNG. ConclusionsVestibular dysfunction may be a common finding in children with EVA. Clinicians who provide medical care for children with EVA need to be familiar with signs of potential balance and vestibular impairments. Although performing vestibular evaluation on young children with EVA can be difficult, objective testing is important in order to identify any potential vestibular deficit in these pediatric patients so that proper vestibular rehabilitation and balance retraining can be provided.

On the deficiency of intelligibility metrics as proxies for subjective intelligibility

A recent trend in deep neural network (DNN)-based speech enhancement consists of using intelligibility and quality metrics as loss functions for model training with the aim of achieving high subjective speech intelligibility and perceptual quality in real-life conditions. In this study, we analyze a variety of loss functions, including some based on state-of-the-art intelligibility and quality metrics, to train an end-to-end speech enhancement system based on a fully convolutional neural network. The loss functions include perceptual metric for speech quality evaluation (PMSQE), scale-invariant signal-to-distortion ratio (SI-SDR), SI-SDR integrating speech pre-emphasis, short-time objective intelligibility (STOI), extended STOI (ESTOI), spectro-temporal glimpsing index (STGI), and a composite loss function combining STGI and SI-SDR. While DNNs trained with these loss functions produce notable speech intelligibility (and quality) gains according to pertinent objective metrics, we conduct a subjective intelligibility test that contradicts this result, showing no intelligibility improvement. From the results of this study, our conclusion is twofold: (1) subjective intelligibility evaluation is currently not replaceable by objective intelligibility evaluation, and (2) both the development of meaningful intelligibility metrics and DNN-based speech enhancement systems that can consistently improve the intelligibility of noisy speech for human listening remain open problems.

MUC5B Alters Oral Mucosa of Oral Lichen Planus Patients with Xerostomia

Oral Lichen planus（OLP）is a chronic autoimmune cell-mediated inflammatory disease. Most OLP patients complain of xerostomia/dry mouth due to changes in the oral mucosal protective barrier. Mucin 5B（MUC5B）is the predominate gel-forming mucin in the oral cavity that protects the underlying tissue from mechanical damage, chemical insult, and pathogenic microbes. This study aims to assess MUC5B in OLP subjects and correlate its role in etiopathogenesis. The study included 20 OLP subjects and 10 healthy controls. OLP subjects were divided further into 10 keratotic（reticular）and 10 non-keratotic（atrophic/erythematous）groups based on the clinical criteria. Dryness of mouth was confirmed by the subjective（Xerostomia Inventory, XI; Bother Index, BI）and clinical testing（Clinical Oral Dryness Score, CODS）method. Stimulated, unstimulated saliva and blood samples were collected from subjects. Samples were centrifuged and analysed for MUC5B level using an ELISA kit. The subjective and clinical testing methods which clinically define xerostomia are correlated with the age and sex of the OLP subjects. The stimulated and unstimulated salivary flow rate of OLP subjects was lower（p<0.05）than controls and was associated with the XI, BI, and CODS（p<0.01）. In contrast, the MUC5B level was increased in the serum of OLP subjects compared to controls（p<0.05）. OLP subjects with xerostomia, displayed lower salivary flow, decrease MUC5B in saliva and increase MUC5B in serum. Alteration of MUC5B in the saliva causes oral mucosal alterations affecting the protection of the oral cavity and increase xerostomia in OLP.

Analisis Kelelahan Kerja Subjektif dengan Menggunakan Kuesioner Subjective Self Rating Test

Occupational health and safety need to be carried out properly in a job as an effort to reduce the occurrence of work fatigue. Work fatigue is a condition in which a person experiences both physical and mental deterioration . When someone experiences work fatigue, his work productivity, efficiency and effectiveness can also decrease. Excessive working time from existing rules can result in work fatigue . PT. United Steel Center has a division with a very long working time duration of 33,420 minutes or 557 hours in February 2022. Based on this, it is necessary to conduct an analysis to find out whether there is subjective fatigue in machine operators in the division. This study uses the method of Subjective Self Rating Test (SSRT). Based on the results of distributing the SSRT questionnaire and analysis, it is known that from eight operators , there are seven operators who experience low work fatigue, while one operator experiences high work fatigue.

Comparison of deep learning models for digital H&E staining from unpaired label-free multispectral microscopy images

Background and objective: This paper presents the quantitative comparison of three generative models of digital staining, also known as virtual staining, in H&E modality (i.e., Hematoxylin and Eosin) that are applied to 5 types of breast tissue. Moreover, a qualitative evaluation of the results achieved with the best model was carried out. This process is based on images of samples without staining captured by a multispectral microscope with previous dimensional reduction to three channels in the RGB range.Methods: The models compared are based on conditional GAN (pix2pix) which uses images aligned with/without staining, and two models that do not require image alignment, Cycle GAN (cycleGAN) and contrastive learning-based model (CUT). These models are compared based on the structural similarity and chromatic discrepancy between samples with chemical staining and their corresponding ones with digital staining. The correspondence between images is achieved after the chemical staining images are subjected to digital unstaining by means of a model obtained to guarantee the cyclic consistency of the generative models.Results: The comparison of the three models corroborates the visual evaluation of the results showing the superiority of cycleGAN both for its larger structural similarity with respect to chemical staining (mean value of SSIM ∼ 0.95) and lower chromatic discrepancy (10%). To this end, quantization and calculation of EMD (Earth Mover’s Distance) between clusters is used. In addition, quality evaluation through subjective psychophysical tests with three experts was carried out to evaluate quality of the results with the best model (cycleGAN).Conclusions: The results can be satisfactorily evaluated by metrics that use as reference image a chemically stained sample and the digital staining images of the reference sample with prior digital unstaining. These metrics demonstrate that generative staining models that guarantee cyclic consistency provide the closest results to chemical H&E staining that also is consistent with the result of qualitative evaluation by experts.

Deep learning-based speaker-adaptive postfiltering with limited adaptation data for embedded text-to-speech synthesis systems

End-to-end (e2e) speech synthesis systems have become popular with the recent introduction of text-to-spectrogram conversion systems, such as Tacotron, that use encoder–decoder-based neural architectures. Even though those sequence-to-sequence systems can produce mel-spectrograms from the letters without a text processing frontend, they require substantial amounts of well-manipulated, labeled audio data that have high SNR and minimum amounts of artifacts. These data requirements make it difficult to build end-to-end systems from scratch, especially for low-resource languages. Moreover, most of the e2e systems are not designed for devices with tiny memory and CPU resources. Here, we investigate using a traditional deep neural network (DNN) for acoustic modeling together with a postfilter that improves the speech features produced by the network. The proposed architectures were trained with the relatively noisy, multi-speaker, Wall Street Journal (WSJ) database and tested with unseen speakers. The thin postfilter layer was adapted with minimal data to the target speaker for testing. We investigated several postfilter architectures and compared them with both objective and subjective tests. Fully-connected and transformer-based architectures performed the best in subjective tests. The novel adversarial transformer-based architecture with adaptive discriminator loss performed the best in the objective tests. Moreover, it was faster than the other architectures both in training and inference. Thus, our proposed lightweight transformer-based postfilter architecture significantly improved speech quality and efficiently adapted to new speakers with few shots of data and a hundred training iterations, making it computationally efficient and suitable for scalability.

Serum Vitamin D Levels and Dry Eye Disease in Postmenopausal Women: A Case-Control Study at a Tertiary Care Center in Rural Haryana.

Despite the high prevalence of Vitamin-D insufficiency and high susceptibility to dry eye disease (DED) in postmenopausal women (PMW), correlation between DED and Vitamin D has not been explored in PMW in any Indian study. To explore the correlation between serum Vitamin D levels in PMW with and without DED, in a hospital-based population in rural Haryana. Subjective (ocular surface disease index [OSDI] questionnaire) and objective clinical tests were undertaken for DED diagnosis. 25(OH) Vitamin D was measured in serum using enzyme-linked immunosorbent assay; insufficient (10-30 ng/ml) and deficient (<10 ng/ml). Descriptive statistics were analyzed by mean ± standard deviation for continuous and frequencies for the categorical variables; Student's t-test used to find out mean difference in Vitamin D levels; P < 0.05 was considered statistically significant. One hundred and forty PMW (60.1 ± 5.32 years) were included; Group-A (Controls; no DED; n: 70); Group-B (Cases; DED diagnosed by OSDI scores; n: 70); Subgroup-B1 (clinical tests negative; n: 30) and B2 (clinical tests positive; n: 40). There was no statistically significant difference in OSDI scores between B1 and B2. Significantly lower mean Vitamin D levels were found in cases (14.36 ± 4.08 ng/ml) as compared to controls (19.19 ± 6.4 ng/ml) (P = 0.001) and in B2 (13.15 ± 3.51 ng/ml) as compared to B1 (15.57 ± 4.66 ng/ml) (P = 0.01). There were significantly low levels of Vitamin-D in clinically established DED. Evaluating Vitamin D levels as a part of the dry eye workup in PMW is recommended. OSDI scores were not aligned with the clinical test scores; questionnaire-based tests alone may not be sufficient for diagnosing DED.

Binocular Rivalry Impact on Macroblock-Loss Error Concealment for Stereoscopic 3D Video Transmission

Three-dimensional video services delivered through wireless communication channels have to deal with numerous challenges due to the limitations of both the transmission channel's bandwidth and receiving devices. Adverse channel conditions, delays, or jitters can result in bit errors and packet losses, which can alter the appearance of stereoscopic 3D (S3D) video. Due to the perception of dissimilar patterns by the two human eyes, they can not be fused into a stable composite pattern in the brain and hence try to dominate by suppressing each other. Thus, a psychovisual sensation that is called binocular rivalry occurs. As a result, undetectable changes causing irritating flickering effects are seen, leading to visual discomforts such as eye strain, headache, nausea, and weariness. This study addresses the observer's quality of experience (QoE) by analyzing the binocular rivalry impact on the macroblock (MB) losses in a frame and its error propagation due to predictive frame encoding in stereoscopic video transmission systems. To simulate the processing of experimental videos, the Joint Test Model (JM) reference software has been used as it is recommended by the International Telecommunication Union (ITU). Existing error concealing techniques were then applied to the contiguous lost MBs for a variety of transmission impairments. In order to validate the authenticity of the simulated packet loss environment, several objective evaluations were carried out. Standard numbers of subjects were then engaged in the subjective testing of common 3D video sequences. The results were then statistically examined using a standard Student's t-test, allowing the impact of binocular rivalry to be compared to that of a non-rivalry error condition. The major goal is to assure error-free video communication by minimizing the negative impacts of binocular rivalry and boosting the ability to efficiently integrate 3D video material to improve viewers' overall QoE.

Evaluation of the Mental Workload of PSIT Employees at SIT XYZ Institutions

The results of the Subjective Self Rating Test (SSRT) questionnaire in the Information and Technology System Development section showed that workers experienced symptoms of severe head, fatigue, drowsiness, yawning, forgetfulness, awkwardness and stiffness. This indicates a problematic workload. This study aims to see the level of mental workload of each staff. The method in this study used NASA-TLX. Based on calculations, it was found that staff 4 has a mental workload score of 84.67 which means it is very high, staff 6 has a mental workload score of 27.33 with a low category, and for other staff has a high mental workload, which is a range of values from 66 to 74.67.

Production of hydrophilic silicone softener with antibacterial properties and evaluation of its performance on fabrics

The goal of the present paper is to produce antimicrobial textile softeners by using a methanol extract of the bitter melon. For this purpose, hydrophilic silicone softeners, including bitter melon extract were prepared at different concentrations. The antibacterial silicone softener with the highest antibacterial activity was determined through a quantitative antibacterial test method (disc diffusion and AATCC 147). The antibacterial silicone softener was applied to 100% cotton knitted fabrics, and the fabric performances were investigated via hydrophilicity tests, bursting strength, bending rigidity, subjective handle tests, whiteness degree and CIEL*a*b* color values. Furthermore, the antibacterial activities of the fabrics were quantitively researched by using ATTCC 100 test method according to S. aureus and E. coli bacteria species. The results showed that it is possible to produce antibacterial silicone softeners by using methanol extract of the bitter melon, and the fabric samples gained antibacterial activity without adversely affecting their performance. The fabrics treated with antibacterial silicone softener killed 65.55% and 52.35% of bacteria after 24 h for S. aureus and E.coli, respectively.

In childhood, a different electrode placement improves the recording of pattern reversal visual evoked potentials.

In childhood, a different electrode placement improves the recording of pattern reversal visual evoked potentials.

PENGARUH MODEL PEMBELAJARAN CONCEPT SENTENCES UNTUK MENINGKATKAN KETERAMPILAN MENULISTEKS DESKRIPSI PADA SISWA KELAS X SMA NURUL IMAN TANJUNG MORAWA TAHUN PEMBELAJARAN 2021-2022

This study aims to determine the effect of the Contextual Teaching and Learning Learning Model on the Ability to Write News Texts for Grade VIII students of MTs Darul Arifin in the 2021-2022 Academic Year. The method used in this research is quasi-experimental. Sampling was carried out using a purposive sampling technique. The research sample consisted of 20 experimental and 20 control class students. Measurement of student learning outcomes based on research instruments in the form of subjective tests in the form of essays. Based on the results of data analysis, it can be concluded that there is a significant influence on learning outcomes in Writing News Texts using the Contextual Teaching and Learning Learning Model. This is shown from the results of the calculation of the posttest hypothesis test through the t-test at a significant level of 0.05, where tcount&gt;ttable (2.864154 &gt; 1.721296296). The achievement of the average student score in the Contextual Teaching and Learning learning class with a value of 84.25 reinforces this result. In contrast, conventional learning obtains a value of 79.5. These results indicate that student activities in the Contextual Teaching and Learning learning model are better.

Acoustic Change Complex as a Neurophysiological Tool to Assess Auditory Discrimination Skill: A Review.

Introduction Acoustic change complex (ACC) is a type of event-related potential evoked in response to subtle change(s) in the continuing stimuli. In the presence of a growing number of investigations on ACC, there is a need to review the various methodologies, findings, clinical utilities, and conclusions of different studies by authors who have studied ACC. Objective The present review article is focused on the literature related to the utility of ACC as a tool to assess the auditory discrimination skill in different populations. Data Synthesis Various database providers, such as Medline, Pubmed, Google, and Google Scholar, were searched for any ACC-related reference. A total of 102 research papers were initially obtained using descriptors such as acoustic change complex , clinical utility of ACC , ACC in children , ACC in cochlear implant users , and ACC in hearing loss . The titles, authors, and year of publication were examined, and the duplicates were eliminated. A total of 31 research papers were found on ACC and were incorporated in the present review. The findings of these 31 articles were reviewed and have been reported in the present article. Conclusion The present review showed the utility of ACC as an objective tool to support various subjective tests in audiology.

Multidimensional evaluation of teaching strategies for pharmacology based on a comprehensive analysis involving 21,269 students.

Background: Given the limitations of traditional pharmacology pedagogical method, diverse novel teaching methods have been widely explored. In this study, we performed a network meta-analysis (NMA) to evaluate the effects of different strategies in pharmacology education. Methods: Literature databases were searched from their inception to November 2022, and the studies were screened according to predefined inclusion and exclusion criteria to extract important information. Outcomes, including theoretical test scores, experimental test scores, subjective test scores, satisfaction scores, and the proportion of satisfaction, were analyzed using R software (version 3.6.1) and STATA (version 15). The NMA was conducted with a random-effects model under the Bayesian framework to calculate odds ratios (ORs) or mean differences (MDs) with associated 95% credible intervals (95% CIs). Surface under the cumulative ranking curve (SUCRA) probability values were calculated to rank the teaching methods examined. Results: A total of 150 studies involving 21,269 students were included. This NMA systematically evaluated 24 teaching strategies, such as problem-based learning (PBL), team-based learning (TBL), case-based learning (CBL) and flipped classrooms (FC), etc., The results of the NMA showed that, PBL combined with CBL was most likely to improve students' theoretical and subjective test scores (SUCRA = 75.49 and 98.19%, respectively), TBL was most likely to improve the experimental test score (SUCRA = 92.38%) and the satisfaction score (SUCRA = 88.37%), while FC had the highest probability of being the best option for improving the proportion of satisfaction (SUCRA = 84.45%). Conclusion: The current evidence indicates that TBL, PBL combined with CBL, and FC might be optimal strategies for pharmacology education since they have a more beneficial effect on students.