Bundled or unbundled? A multi-text corpus-assisted discourse analysis of the relationship between teaching and research in UK universities

Creating and disseminating knowledge through research and teaching has long been regarded as the hallmark of the modern university. However, new university business models have called into question the ‘bundling’ of teaching and research, and sustained research on the relationship between teaching and research has found little evidence of an insoluble connection between the two activities. In this article, we explore the relationship between teaching and research from the perspective of universities’ institutional discourse. We use corpus-assisted discourse analysis to examine the relationship between research and teaching as presented in two sets of institutional texts currently influential in UK Higher Education: Research Excellence Framework environment statements and Teaching Excellence Framework provider submissions (a total of 2143 documents and 12,492,071 words). Our findings show that, while universities emphasise the value of research to their teaching, they do not always emphasise (or sometimes even decry) the influence of teaching on their research. We empirically evidence that, according to what universities themselves write in institutional texts, teaching and research are not always in a mutually beneficial entanglement, but often rather a one-way relationship in which research expertise and institutional prestige are used to bolster claims of teaching excellence. This has implications for the communication of both the vision and the purpose of a university in regulatory exercises and wider policy, but also speaks to the broader idea and practice of being a university in the twenty-first century.     

专利被引频次的时间影响研究

[目的/意义]研究时间因素对专利被引频次的影响,可以减少时间因素对技术评价活动的制约,提高评价的准确性和可信度。[方法/过程]采集1975-2017年的美国专利数据,开展基于固定效应法的专利被引频次的修正研究。将专利按照不同公开年份和不同技术领域分组,选定组内均值和6个TOP分位数为被引频次基准,统计当前时间点的被引频次基准线及基准线的历史时序变化情况。建立神经网络模型,拟合基准线的时序变化规律,并预测未来统计时间点的基准线。[结果/结论]专利公开年份和统计年份的时间差异,使得专利被引频次无法直接进行比较。本文建立了基于不同技术领域、不同公开年份和不同统计年份的专利被引频次基准线,为专利评估提供参考。     

动态助词“来”的历时发展及其衰落动因考察

古代汉语中,"来"的用法异常复杂,仅其助词属性下就又可以细分出六七种不同的用法,但这些用法大多并未能沿用到现代汉语共同语中,这也包括用作动态助词的"来"。从历时发展来看,动态助词"来"产生之后,在唐五代、宋元、明清三个阶段的使用并未呈现出持续上升的态势,而是在共同语中逐渐衰落了,今天仅在偏南方的一些方言中还有所使用。作为动态助词,"来"的衰落动因是多方面的,既有"来"自身功能负担过重,以及在使用中受限制过多的内因作用,也有系统内部成员竞争的外因作用。     

Does modality matter? A latent profile and transition analysis of sedentary behaviours among school-aged youth: The UP&DOWN study

ABSTRACT

The aims of this study were to identify profiles of sedentary behaviour (SB) patterns, based on leisure-time self-reported SB modalities (screen, educative, social, and relaxing) and to evaluate changes in these profiles over 2 years among Spanish youth aged 8–18 years. Latent profile analysis (LPA), a data-driven analytic approach, was used to identify groups of boys and girls (n = 1553; 48% girls; mean±SD age: 12.56 ± 2.49 y) with distinct SB profiles using the SB modalities (time/d) as input variables. Latent transition analysis, an extension of LPA that uses longitudinal data, was used to analyse 2-year changes in these profiles. At baseline, four and three SB profiles were found among boys (labelled: screen, educative, social, and relaxing) and girls (labelled: screen/social, educative, and relaxing), respectively. Overall, more girls (range: 48%-67%) had the same profile over time, than boys (40%-52%). Participants with a screen or relaxing SB profile at baseline were more likely to have an educative profile after 2 years. Youth with a social and an educative SB profile at baseline were more likely to transition to profiles characterized by higher screen and social SB, respectively. Using a novel and person-centered approach, this study identified gender-specific SB profiles that were moderately stable over time.     

Syllable-based reading improvement: Effects on word reading and reading comprehension in Grade 2

This study examined the effects of a syllable-based reading intervention for German second graders who demonstrated difficulties in the recognition of written words. The intervention focused on fostering word reading via syllable segmentation. The materials consisted of the 500 most frequent syllables typically read by 6- to 8-year-old children. The aims were to practice phonological recoding, consolidate orthographic representations of syllables, and routinize the access to these representations. Compared to children randomly assigned to a wait-list group, poor readers in the treatment condition showed significant improvements in standardized measures of phonological recoding, direct word recognition, and text-based reading comprehension after the 24-session intervention. Poor readers in the treatment condition also showed greater improvements in development of word recognition compared to children with efficient word recognition skills. The results provide evidence that a syllable-based reading intervention is a promising approach to increase struggling readers’ word recognition skills, which in turn will improve their reading comprehension.     

基于深度学习的中文专利自动分类方法研究

[目的/意义] 面向当前国内专利审查和专利情报分析工作中对于海量专利分类的客观需求，设计了7种基于深度学习的专利自动分类方法，对比各种方法的分类效果，从而助力专利分类效率和效果的提升。[方法/过程] 针对传统机器学习方法存在的缺陷，基于Word2Vec、CNN、RNN、Attention机制等深度学习技术，考虑专利文本语序特征、上下文特征以及分类关键特征，设计Word2Vec+TextCNN、Word2Vec+GRU、Word2Vec+BiGRU、Word2Vec+BiGRU+TextCNN等7种深度学习模型，以中国专利为例，选取IPC主分类号的"部"作为分类依据，对比这7种模型与3种传统分类模型在中文专利分类任务中的效果。[结果/结论] 实证研究效果显示，采用考虑语序特征、上下文特征及强化关键特征的深度学习方法进行中文专利分类具有更优的分类效果。     

图书情报学领域学科交叉及其国家差异的动态分析

[目的/意义]领域知识的跨学科交叉研究能够打破学科间的壁垒,有助于发现重大科学问题的解决方案。[研究设计/方法]基于图书情报学领域文献题录信息构建轻量级领域知识图谱,从中提取学科信息、国家信息、时间信息及其关联,采用时间与空间相结合的多维度分析方法,对学科交叉的演化进程以及国家差异进行跟踪与分析。[结论/发现]图书情报学领域内学科交叉现象日渐显著,各国家的学科交叉程度与倾向存在差异,一些目前尚不突出的交叉学科方向值得关注。[创新/价值]采用多维度视角分析学科交叉现象,相关结论可以为国家科技战略制定及学科发展规划提供有益参考。     

Olfaction and emotion: The quest for olfactory restoration in two speech communities

ABSTRACT

Olfaction has been described as one of the least studied of the senses. This has certainly proven true in Communication. This project applies Ethnography of Communication and Cultural Discourse Analysis to examine the role that olfaction plays in the cultural meaning systems of two speech communities: residents surrounding the Gowanus Canal and bereaved parents. The cultural hub used to explore the cultural significance of olfaction in both speech communities is feeling. The analyses illustrate how emotions are tied to associated sensations and further become activated through radiants of dwelling, relating and acting. The analysis also shows how olfaction may serve to unify and separate members of a community. The significance of this research encourages stretching the boundaries of communication studies by examining the less studied physical senses and the boundaries of Cultural Discourse Analysis by providing distinctions that others can incorporate when examining feelings.     

Comparing cumulative index and factor analytic approaches to measuring maltreatment in the National Longitudinal Study of Adolescent to Adult Health

Child maltreatment is a complex and multifaceted construct in need of advanced statistical techniques to improve its measurement. The current study compared the predictive utility of a cumulative index to a factor analytic approach for constructing a measure of maltreatment. Data were from Waves III and IV of the National Longitudinal Study of Adolescent to Adult Health (Wave III: n = 14,800; Wave IV: n = 12,288). As adults, participants retrospectively reported on their childhood experiences of physical abuse, sexual abuse, emotional abuse, physical neglect, supervisory neglect, and social services investigations. Both the cumulative index and a two-factor solution showed evidence of convergent validity, predicting lifetime incidence of homelessness, being paid for sex, and various measures of running away or living apart from biological parents, and prospectively predicting depression, substance use, and criminal behavior. The latent variables, derived from a factor analytic approach, had greater explanatory power for many outcomes compared to the cumulative index, even when controlling for sociodemographic variables. Results suggest that factor analysis is a better methodology than a cumulative index for measuring maltreatment in large datasets when explanatory power for external outcomes is of greatest concern.     

Digital Module 16: Longitudinal Data Analysis

In this digital ITEMS module, Dr. Jeffrey Harring and Ms. Tessa Johnson introduce the linear mixed effects (LME) model as a flexible general framework for simultaneously modeling continuous repeated measures data with a scientifically defensible function that adequately summarizes both individual change as well as the average response. The module begins with a nontechnical overview of longitudinal data analyses drawing distinctions with cross-sectional analyses in terms of research questions to be addressed. Nuances of longitudinal designs, timing of measurements, and the real possibility of missing data are then discussed. The three interconnected components of the LME model—(1) a model for individual and mean response profiles, (2) a model to characterize the covariation among the time-specific residuals, and (3) a set of models that summarize the extent that individual coefficients vary—are discussed in the context of the set of activities comprising an analysis. Finally, they demonstrate how to estimate the linear mixed effects model within an open-source environment (R). The digital module contains sample R code, diagnostic quiz questions, hands-on activities in R, curated resources, and a glossary.