Working while studying: the impact of term-time employment on dropout rates

There are many possible reasons why students leave university prior to degree completion, and one of the more commonly cited is being employed while studying. This paper analyses the impact of employment status on dropout rates using survival analysis. It finds that employment status does have an impact on dropout rates; students who work full time alongside studying full time are less likely to complete their programme than students working short part-time or not working at all. However, it seems as if there is a threshold to how much students can work, as working more than 20?h a week (long part-time work) increase the risk of dropout as much as full-time work. Integrating employment status into the analysis does not change the effect of variables known to have an influence on dropout, such as grades, gender and social background, but it contributes to further explain who are at risk of dropout. This implies that models for dropout and retention must also take such external factors into account, not just consider what happens at university, as in model of student departure.     

Student use of out-of-class study groups in an introductory undergraduate biology course

Self-formed out-of-class study groups may benefit student learning; however, few researchers have quantified the relationship between study group use and achievement or described changes in study group usage patterns over a semester. We related study group use to performance on content exams, explored patterns of study group use, and qualitatively described student perceptions of study groups. A pre- and posttest were used to measure student content knowledge. Internet-based surveys were used to collect quantitative data on exam performance and qualitative data on study group usage trends and student perceptions of study groups. No relationship was found between gains in content knowledge and study group use. Students who participated in study groups did, however, believe they were beneficial. Four patterns of study group use were identified: students either always (14%) or never (55%) used study groups, tried but quit using them (22%), or utilized study groups only late in the semester (9%). Thematic analysis revealed preconceptions and in-class experiences influence student decisions to utilize study groups. We conclude that students require guidance in the successful use of study groups. Instructors can help students maximize study group success by making students aware of potential group composition problems, helping students choose group members who are compatible, and providing students materials on which to focus their study efforts.     

Psycho-Education with Asylum Seekers and Survivors of Torture

This article examines psycho-educational programmes for asylum-seekers and tortured refugees at the Danish Red Cross Asylum Department and Rehabilitation and Research Centre for Torture victims respectively. The psycho-education programme is based on a cognitive theoretical framework. However, it is argued that the processes observed during the programmes and the changes in the participants’ lives may be conjointly understood within theoretical frameworks of narrative therapy, social constructionism, and community psychology, emphasizing professionals’ and participants’ co-construction of alternative stories and action possibilities, and the importance of the context and social network outside the intervention context.

Pedagogical possibilities for unruly bodies

Project Re?Vision uses disability arts to disrupt stereotypical understandings of disability and difference that create barriers to healthcare. In this paper, we examine how digital stories produced through Re?Vision disrupt biopedagogies by working as body-becoming pedagogies to create non-didactic possibilities for living in/with difference. We engage in meaning making about eight stories made by women and trans people living with disabilities and differences, with our interpretations guided by the following considerations: what these stories ‘teach’ about new ways of living with disability; how these stories resist neoliberalism through their production of new possibilities for living; how digital stories wrestle with representing disability in a culture in which disabled bodies are on display or hidden away; how vulnerability and receptivity become ‘conditions of possibility’ for the embodiments represented in digital stories; and how curatorial practice allows disability-identified artists to explore possibilities of ‘looking back’ at ableist gazes.     

Enhancement of biosensing performance in a droplet-based bioreactor by in situ microstreaming

A droplet-based micro-total-analysis system involving biosensor performance enhancement by integrated surface-acoustic-wave (SAW) microstreaming is shown. The bioreactor consists of an encapsulated droplet with a biosensor on its periphery, with in situ streaming induced by SAW. This paper highlights the characterization by particle image tracking of the speed distribution inside the droplet. The analyte-biosensor interaction is then evaluated by finite element simulation with different streaming conditions. Calculation of the biosensing enhancement shows an optimum in the biosensor response. These results confirm that the evaluation of the Damköhler and Peclet numbers is of primary importance when designing biosensors enhanced by streaming.It has been pointed out that biosensing performances can be limited by the diffusion of the analytes near the sensing surface.¹ In the case of low Peclet number hydrodynamic flows, typical of microfluidic systems, molecule displacements are mainly governed by diffusive effects that affect time scales and sensitivity. To overcome this problem, the enhancement of biosensor performance by electrothermal stirring within microchannels was first reported by Meinhart et al.² Other authors^{3, 4} numerically studied the analyte transport as a function of the position of a nanowire-based sensor inside a microchannel, stressing on the fact that the challenge for nanobiosensors is not the sensor itself but the fluidic system that delivers the sample. Addressing this problem, Squires et al.⁵ developed a simple model applicable to biosensors embedded in microchannels. However, the presented model is limited to the case of a steady flow. The use of surface-acoustic waves (SAWs) for stirring in biomicrofluidic and chemical systems is becoming a popular investigation field,^{6, 7, 8, 9} especially to overcome problems linked to steady flows by enhancing the liquid∕surface interaction.^{1, 10, 11} The main challenges that need to be addressed when using SAW-induced stirring are the complexity of the flow and its poor reproducibility. However, some technical solutions were proposed to yield a simplified microstreaming. Yeo et al. presented a centrifugation system based on SAW that produces the rotation of the liquid in a droplet in a reproducible way by playing on the configuration of the transducers and reflectors,¹² and presented a comprehensive experimental study of the three-dimensional (3D) flow that causes particle concentration in SAW-stirred droplets,¹³ revealing the presence of an azimuthal secondary flow in addition to the main vortexlike circular flow present in acoustically stirred droplets. The efficiency of SAW stirring in microdroplets to favorably cope with mass transport issues was finally shown by Galopin et al.,¹⁴ but the effect of the stirring on the analyte∕biosensor interaction was not studied. It is expected to overcome mass transport limitations by bringing fresh analytes from the bulk solution to the sensing surface.The studied system, described in Fig. ​Fig.1,1, consists of a microliter droplet microchamber squeezed between a hydrophobic piezoelectric substrate and a hydrophobic glass cover. Rayleigh SAWs are generated using interdigitated transducers (interdigital spacing of 50 μm) laid on an X-cut LiNbO₃ substrate.^{1, 15, 16} The hydrophobicity of the substrate and the cover are obtained by grafting octadecyltrichlorosilane (OTS) self-assembled monolayers (contact angle of 108° and hysteresis of 9°). To do so, the surface is first hydroxylized using oxygen plasma (150 W, 100 mT, and 30 sccm³ O₂) during 1 min and then immersed for 3 h into a 1 mM OTS solution with n-hexane as a solvent.Open in a separate windowFigure 1(a) General view of the considered system. (b) Mean value of the measured speeds within the droplet as a function of the inlet power before amplification.When Rayleigh waves are radiated toward one-half of the microchamber, a vortex is created in the liquid around an axis orthogonal to the substrate due to the momentum transfer between the solid and the liquid. This wave is generated under the Rayleigh angle into the liquid.Speed cartographies of the flow induced in the droplet are realized using the particle image tracking technique for different SAW generation powers. To do so, instantaneous images of the flow are taken with a high-speed video camera at 200 frames∕s and an aperture time of 500 μs on a 0.25 μl droplet containing 1 μm diameter fluorescent particles. Figure ​Figure11 shows the mean speed measured in the droplet as a function of the inlet power. The great dependence of the induced mean speed with the SAW power enables a large range of flow speeds in the stirred droplet. Moreover, the flow was visualized with a low depth of field objective. It was found to be circular and two dimensional (2D) in a large thickness range of the droplet.The binding of analytes to immobilized ligands on a biosensor is a two step process, including the mass transport of the analyte to the surface, followed by a complexation step,Abulk↔kmAsurface+B↔ka,kdAB(1)with k_m as the constant rate for mass transport from and to the sensor, and k_a and k_d as the constant rates of association and dissociation of the complex.At the biosensor surface, the reaction kinetics consumes analytes but their transport is limited by diffusive effects. In this case, the Damköhler number brings valuable information by comparing these two effects. Calling the characteristic time of reaction and diffusion, respectively, τ_C and τ_M, the mixing time in diffusion regime can be approximated by τ_M≈h²∕D with D as the diffusion coefficient and h a characteristic length of the microchannel. Calling R_T the ligand concentration on the surface in mole∕m², the Damköhler number (D_a) can be written asDa=τMτC=kaRThD.(2)Depending on the type of reaction, the calculation of D_a helps determine if a specific biointeraction will benefit from a mass SAW-based microstreaming. If the Damköhler number is low, the reaction is slow compared to mass transport and the reaction will not significantly benefit from microstirring. For example, the hybridization of 19 base single stranded DNA in a microfluidic system with a characteristic length of 500 μm is characterized by a Damköhler number of 0.07 and is therefore not significantly influenced by mass transport. On the contrary, the binding of biotin to immobilized streptavidin is characterized by a D_a number of approximately 10⁴. In this case, the stirring solution will significantly improve the reaction rate.COMSOL numerical simulations were carried out to study the efficiency of the SAW stirring in the case of a droplet-based microbioreactor with a diameter of 1 mm. Assuming a 2D flow, the simulated model takes into account the convective and diffusive effects in the analyte-carrying fluid and the binding kinetics on the biosensor surface. This approach was thoroughly developed by Meinhart et al.²On the biosensor surface, the following equations are solved:∂B∂t=kacs(RT−B)−kdB,(3)∂B∂t=D|∂c∂y|y=0(4)with c as the local concentration of analytes in the droplet and B as the surface concentration of bound analytes on the biosensor surface. Simulation results show that a depleted zone is formed near the biosensor in the case of an interaction without stirring. This zone is characterized by a low concentration of analytes and results from the trapping of analytes on the biosensor surface, thus creating a concentration gradient on the vicinity of the biosensor. When stirring is applied, the geometry of the depleted zone is modified, as it is pushed in the direction of the flow. The geometry of the depleted zone then depends on many parameters, among which the diffusion coefficient D, the speed distribution of the flow (not only near the biosensor but also in the whole microfluidic system), and the reaction kinetics on the biosensor. In our case, which is assimilated to a simple circular flow, the depleted zone reaches a permanent state consisting of an analyte-poor layer situated in the exterior perimeter of the stirred droplet. The diffusion of analytes is then limited again by diffusion from the inner part of the droplet toward its exterior perimeter (see Fig. ​Fig.22).Open in a separate windowFigure 2(a) Mean concentration of bound analytes vs time for different mean flow speeds. (b) The obtained concentration profiles with and without circular stirring, t=10 000 s.The initial analyte and receptor concentrations are, respectively, 0.1 nM in the solution and 3.3×10⁻³ nM m on the biosensor surface, the diffusion coefficient is D=10⁻¹¹ m² s⁻¹, and the reaction constants are k_a=10⁶ M⁻¹ s⁻¹ and k_d=10⁻³ s⁻¹. Simulations show that the mean concentration of bound analytes highly increases with the flow speed, improving the efficiency of the biosensing device. To evaluate the benefits of in situ microstreaming with SAW, the same simulations were conducted for D_a numbers ranging from 10⁴ to 10⁸ M⁻¹∕s, by ranging the diffusion coefficient from 4×10⁻¹² to 4×10⁻⁹ m²∕s, and the association coefficient k_a from 10⁴ to 10⁸ M⁻¹∕s. The enhancement factor of analyte capture, defined as the ratio of the binding rate with streaming B and the binding rate without streaming B0, is plotted in Fig. ​Fig.33 for different values of D_a. Calculations are done in the case of a mean flow speed of 0.5 mm∕s.Open in a separate windowFigure 3(a) Enhancement factor (defined as the ratio between binding rate with streaming B and binding rate without streaming B0) for different Damkhöler numbers and (b) normalized enhancement factor for different Peclet numbers.One can notice the saturation of the enhancement factor curve for large value of D_a to the value of 3.5 for high D_a. This can be explained by the fact that for large k_a∕D_a ratios, the analytes, which normally require penetration in the depleted zone by diffusion, do not have time to interact with the biosensor when they pass in the vicinity of its surface. The efficiency of the streaming is then reduced for large values of D_a. In the case of our specific flow configuration, the enhancement factor reaches 3.2 for the interaction of streptavidin on immobilized biotin (D_a=10³).The reported simulation results can be compared to an experimental value obtained using the droplet-based surface plasmon resonance sensor streamed in situ using SAW reported by Yeo et al.¹² By monitoring the streptavidin∕biotin binding interaction on an activated gold slide, they showed that SAW stirring brings an improvement factor of more than 2. This difference can be accounted to the high complexity of the induced 3D flow, which was modeled in a simple manner in our calculations.Other factors must be taken into account when optimizing the improvement factor, such as the flow velocity and the characteristic length of the mixing. To do so, the Peclet number allows the comparison of the convective and diffusive effects.¹⁷ For δC a typical variation in concentration on the distance h, the Peclet number is given byPe=UhD.(5)A significantly high Peclet number causes a decrease in biosensing efficiency as the analytes do not have enough time to interact with the biosensing surface by diffusion through the analyte-poor layer. On the contrary, the case of a low Peclet number corresponds to the diffusion-limited problem. Therefore, for each Damköhler number, there is a Peclet number optimizing this factor. To illustrate this fact, Fig. ​Fig.3b3b shows the calculation of the enhancement factor as a function of the Peclet number for a given D_a.In this paper, we showed that surface loading of typical analytes on a droplet-based biosensor can be highly increased by SAW microstirring. The system permits the enhancement of the biosensing performances by the continuous renewal of the analyte-carrying fluid near the sensing surface. Thanks to mean flow speeds measured up to 1800 μm∕s, the SAW microstreaming can be beneficial to the biosensing of a large range of analyte∕ligand interactions. In addition to the biosensing performance improvement, such a method can be easily integrated in micro-micro-total-analysis systems, which makes it a convenient tool for liquid handling in future biochips.     

Using natural language processing to support peer-feedback in the age of artificial intelligence: A cross-disciplinary framework and a research agenda

Advancements in artificial intelligence are rapidly increasing. The new-generation large language models, such as ChatGPT and GPT-4, bear the potential to transform educational approaches, such as peer-feedback. To investigate peer-feedback at the intersection of natural language processing (NLP) and educational research, this paper suggests a cross-disciplinary framework that aims to facilitate the development of NLP-based adaptive measures for supporting peer-feedback processes in digital learning environments. To conceptualize this process, we introduce a peer-feedback process model, which describes learners' activities and textual products. Further, we introduce a terminological and procedural scheme that facilitates systematically deriving measures to foster the peer-feedback process and how NLP may enhance the adaptivity of such learning support. Building on prior research on education and NLP, we apply this scheme to all learner activities of the peer-feedback process model to exemplify a range of NLP-based adaptive support measures. We also discuss the current challenges and suggest directions for future cross-disciplinary research on the effectiveness and other dimensions of NLP-based adaptive support for peer-feedback. Building on our suggested framework, future research and collaborations at the intersection of education and NLP can innovate peer-feedback in digital learning environments.

Practitioner notes

What is already known about this topic

• There is considerable research in educational science on peer-feedback processes.
• Natural language processing facilitates the analysis of students' textual data.
• There is a lack of systematic orientation regarding which NLP techniques can be applied to which data to effectively support the peer-feedback process.

What this paper adds

• A comprehensive overview model that describes the relevant activities and products in the peer-feedback process.
• A terminological and procedural scheme for designing NLP-based adaptive support measures.
• An application of this scheme to the peer-feedback process results in exemplifying the use cases of how NLP may be employed to support each learner activity during peer-feedback.

Implications for practice and/or policy

• To boost the effectiveness of their peer-feedback scenarios, instructors and instructional designers should identify relevant leverage points, corresponding support measures, adaptation targets and automation goals based on theory and empirical findings.
• Management and IT departments of higher education institutions should strive to provide digital tools based on modern NLP models and integrate them into the respective learning management systems; those tools should help in translating the automation goals requested by their instructors into prediction targets, take relevant data as input and allow for evaluating the predictions.