Symposium Timetable 4th Annual Postgraduate Research Symposium Northampton Suite & B307 B/C 26th March 2014 Time 09.45 – 10.15 10.15 – 10.25

Sessions Registration Networking and Tea & Coffee – Northampton Suite B/C Welcome and Introduction Professor Kenneth Grattan – B307 B/C

10.25 – 11.10

Oral presentations

Paper Session 1

Presentation 1 Room 1 Modelling of non-conductive particles suspended in conductive medium Ekaterina Aristovich School of Engineering and Mathematical Sciences Presentation 1 Room 2 Misinformation effects caused by retroactive product placement Maria Hellenthal School of Arts and Social Sciences

11.10 – 11.25

Presentation 2 Room 1 Design and Modelling Methodology of Large Force Long Lifetime Magnetic Shape Memory Actuators Nikita Gabdullin School of Engineering and Mathematical Sciences Presentation 2 Room 2 SenseCam as a tool for studying memory in healthy ageing Ali Mair School of Arts and Social Sciences

Presentation 3 Room 2 The use of eye-tracking methodology in the research of memory in Autism Spectrum Disorders Melanie Ring School of Arts and Social Sciences

Questions and Answers and summing up

11.25 – 11.40

Tea & Coffee View posters

11.45 - 12.30

Oral presentations

Paper Session 2

Presentation 1 Room1 Relational poems – a means to highlight the polyphony within narrative; exploring influence on decision making in young women’s pregnancy stories Sarah Bekaert School of Health Sciences Presentation 1 Room2 Rhythmic abilities of children with language impairments Ioanna Georgiadou School of Health Sciences

12.30 – 12.45

Presentation 3 Room 1 Application of MachineLearning Techniques in Air Transport Judit Guimera-Busquets School of Engineering and Mathematical Sciences

Presentation 2 Room1 Measuring cooperation and food sustainability Raquel Ajates School of Arts and Social Sciences

Presentation 3 Room1 Methodological steps in developing an internet-based anxiety treatment: Current status and future plans Miriam Thiel School of Health Sciences

Presentation 2 Room2 The role of semantically rich gestures in aphasic conversation Judith Kistner School of Health Sciences

Presentation 3 Room2 Let’s move: neurocognitive mechanisms underlying sensorimotor memory Alejandro Galvez-Pol School of Arts and Social Sciences

Questions and Answers and summing up

Room 1 = Northampton Suite A – Chair: Neal Sumner Room 2 = B307 A/B – Chair: Dr Pam Parker

12.45 – 13.45

13.50 – 14.35

Paper Session 3

Lunch Poster Session From 13.15 – 13.45 poster presenters will be by their posters so do go and talk to them about their work and look at the poster. You will be asked to vote for the best poster. Oral Presentations Presentation 1 Room 1 Presentation 2 Room 1 Presentation 3 Room 1 Cooperative Control in Prediction of Turbulent Investigation of thermal effects Networks Diffusion Flames Using on screw compressor Ivana Tomic Tabulated Chemistry clearances School of Engineering and Kai Zhang David Buckney Mathematical Sciences School of Engineering and School of Engineering and Mathematical Sciences Mathematical Sciences Presentation 1 Room 2 Service oriented architectures Shahedeh A.Khani School of Informatics

14.35 – 14.50

Presentation 2 Room 2 Double Reading of Mammograms: Improving Performance by Effectively Pairing Readers and Identifying and Discussing Difficult Cases Marwa Gadala School of Informatics

Questions and Answers and summing up

14.50– 15.05

15.10 – 15.55

Paper Session 4

Tea & Coffee View posters Oral presentations Presentation 1 Room 1 Children’s language skills: are we sure this is what we are talking about? Hannah Hockey School of Health Sciences Presentation 1 Room 2 Moral Dynamics in Everyday Life: How morality evolves in time? Albert Barque-Duran School of Arts and Social Sciences

15.55 – 16.10

Presentation 3 Room 2 Influence of colour and luminance contrast on visual working memory- a study using EEG Maciej Kosilo School of Arts and Social Sciences

Presentation 2 Room 1 What is ‘normal’ in discourse production- and can we measure that? Madeleine Pritchard School of Health Sciences Presentation 2 Room 2 Using Interpretative Phenomenological Analysis to explore the lived experiences of HIV positive mothers in Kingston, Jamaica Gayle Clifford School of Health Sciences

Questions and Answers and summing up

16.10 – 16.20

Closing remarks and award of prizes: Professor Kenneth Grattan

16.20 – 17.00

Tea, Coffee & Cakes

Room 1 = Northampton Suite A – Chair: Neal Sumner Room 2 = B307 A/B – Chair: Dr Pam Parker

Posters will be displayed in Northampton Suite B&C all day. From 13.15 – 13.45 poster presenters will be by their posters so do go and talk to them about their work and look at the poster. You will be asked to vote for best poster.

Poster Presentations 1. Chemotherapy induced cognitive changes in colorectal cancer patients Marie-Rose Dwek School of Health Sciences

2. The Impact of Health Insurance Coverage on Health Care Services among Patients with Chronic Diseases in the Eastern Province of Saudi Arabia Afnan Aljaffary School of Health Sciences

4. The role of image editors in the image selection process in online journalism

5. Measuring the effects of negative emotions on the production of false memories.

Sylwia Frankowska-Takhari School of Informatics

Cassandra Bland School of Arts and Social Sciences

3. Pre-verbal and early verbal communication in siblings of children with language impairment or Autism Spectrum Disorder Helen Cain School of Health Sciences

6. Are these reviews trustworthy? Investigating the role of user interface elements in signalling reviews’ trustworthiness and the effect of user belief Dara Sherwani School of Informatics

7. Measuring Makers Making – Designing an End-User Programming study of Arduino development environments Tracey Booth School of Informatics

8. The use of simulators in the assessment of computational models of learning and behaviour Niklas Kokkola School of Informatics

9. How does the tangible object affect motor skill learning? Milena Markova School of Informatics

Join in on twitter #CityRS2014

We recommend that you ask permission from any identifiable person before posting images or photographs online (e.g. via Twitter)

Book of Abstracts 4th Annual Postgraduate Research Symposium 26th March 2014

Paper Session 1 Presentation 1 Room 1 Modelling of non-conductive particles suspended in conductive medium Ekaterina Aristovich School of Engineering and Mathematical Sciences

The main focus is on investigation of possibility to measure cholesterol non-invasively by electrical impedance technique. This has been done using computational modelling. The physics of conduction is well understood, but the mathematics of describing the current and potential distribution between impedance electrodes, particularly those attached to organic tissue, is more challenging due to non-homogeneity of the system. The proposed modelling methodology is based on minimisation of number of particles and has been successfully used for modelling of cholesterol in human blood. Human blood is a complex conductive medium. Fatty non-conductive substances are suspended in the conductive plasma. From the modelling point of view it is important to note that total cholesterol of 5 mmol/l corresponds to about 1012 lipoprotein particles (5-30 nm in size) in each mm3 of blood. This poses a major challenge for taking into account every individual particle for modelling purposes. This would not only introduce enormous complexity in building finite element (FE) models but also dramatically increase modelling times to an unacceptable level. This problem was tackled by using a particle aggregation strategy in which smaller number of larger particles was used keeping the total volume of non-conductive particles constant. These parameters were then adjusted for FE models in 2D and 3D. However, this simplifying modelling strategy needed to be quantitatively justified by modelling experiments. This was done to establish the maximum number of particles above which no significant change in the modelling results could be obtained by further increasing the number of particles. Furthermore, the effects of randomness in the distribution of these particles on modelling results were quantified. This was done by building numerous models using random number generator for particle placement. It is evidenced from the modelling results this strategy for particle aggregation simplifies models without significantly affecting the results of computation.

Presentation 2 Room 1: Design and Modelling Methodology of Large Force Long Lifetime Magnetic Shape Memory Actuators Nikita Gabdullin School of Engineering and Mathematical Sciences

Magnetic shape memory (MSM) “smart” alloys form a new group of shape memory materials which are very prospective to be used in actuators, sensors and other electrical devices. The main advantage of these alloys is a very high magnetic field induced strain, which is more than 1 order higher than that of giant magnetostrictive materials (GMM). The possibility to control the shape magnetically without any mechanical contact between MSM stick and the core opens up a way to designing very long lifetime devices, in contrast with conventional solenoid-based actuators. However, it is hard to correctly take into account varying properties of MSM element of an actuator. The reason for this is that geometry, resistance and permeability of MSM stick depend on its current strain, whereas twinning stress of a crystal depends on microstructure and is very temperature sensitive. Simple solution of magnetic circuit equations of an actuator can’t give satisfactory results due to complicated relations between different parameters mentioned above. Hence, a new approach for designing an actuator with varying parameters in needed. Modelling of field problems vie finite element method (FEM) allows to significantly increase accuracy of the analysis. On this stage of the research, accurate modelling of electromagnetic field of MSM actuator is the main task. This paper discusses three different methods of modelling the MSM effect in actuators and their compatibility with material’s physics. The first one considers MSM element being always in a single-variant state with its properties depended on strain, the second one considers it being in a two-variant state with volumes of variants depended on strain, and the last one considers

MSM element being in a multi-variant state with a number of variants depended on strain. Modelling results are also compared with analytical circuit-based solutions to illustrate limitations of the latter.

Presentation 3 Room 1: Application of Machine-Learning Techniques in Air Transport Judit Guimera-Busquets School of Engineering and Mathematical Sciences

This research focuses on the application of machine-learning techniques in the air transport sector. Machinelearning is the application of computational methods to identify relationships that exist within large datasets, identifying dominant drivers of important outcomes, predicting the probability of new outcomes occurring, or identifying anomalous behaviour. The main goal of this study will centre on improving and/or developing new features within the existing machine-learning applications in air transport. Although the analysis of how data driven modelling can influence the aviation sector is extensive, my research focuses on two specific areas: aviation safety and air traffic forecasting. The former constitutes the primary objective of this study and its main goal is related to the application of supervised-learning to analyse air safety data provided by an airline, ultimately developing computational tools that can be used by an airline to predict the likelihood of safety incidents, thereby improving airline safety. The latter is centred on applying machinelearning algorithms to predict the future air traffic flows, including predicting the likelihood of airport-pairs previously not connected by flight service, being connected by flight service in the future; and predicting future flight traffic levels on existing and new airport pair services. This will improve current air network forecasts and contribute to improved decision-making by major players in the aviation industry, such as airports and air traffic service providers. Because we are still waiting for air safety data from a partner airline, but already have access to historical air traffic data, this paper will focus on applying supervised-learning to air traffic forecasting, specifically for the domestic air transport system in the United States. Initially, linear regression was applied to identify the significance of different inputs and to identify issues of endogeneity and multi-colinearity. Subsequently, neural networks will be applied to better capture non-linearities in the response of the system. The goal of this research is ultimately to improve and/or develop new features within the existing machine-learning applications in air traffic forecasting, such as extending the prediction of city-pair connections to more detailed traffic information, including flight frequency and aircraft size class.

Presentation 1 Room 2: Misinformation effects caused by retroactive product placement Maria Hellenthal School of Arts and Social Sciences

Research in cognitive psychology has repeatedly shown that misleading information can cause people to falsely remember details which were only suggested to them (e.g. Loftus, 1977). In experimental settings, false memories have been created by various means, the most widely used technique being the misinformation paradigm applied extensively to examine influences on eyewitness testimony (e.g., Loftus, 1996). Here, respondents typically are exposed to events (e.g., a crime scene), receive misinformation afterwards (e.g., that a knife not a gun was used in the scene), and subsequently complete a memory test for the original event. The typical finding is that misinformation is falsely attributed to the original event (misinformation effect) and/or that misinformation leads to an impairment of the original event memory (memory impairment effect). Both phenomena are often explained by source confusions – falsely attributing mental experiences from one source to another source (Brainerd & Reyna, 2005). In this study, a modified version of the misinformation paradigm is applied to examine ‘brand misinformation effects’. More specifically, this research investigates for the first time the effects of product placements in photos as well as the effects of retroactively placed brands in pictures on consumer memory. So far, our studies did not only reveal that product placements in pictures have an influence on adults’ true and false memories, but it was also shown that retroactively placed brands have the potential to impair memory for originally presented brands. However, findings also suggest that the unique characteristics that differentiate brands from critical items that are typically used in misinformation studies have a somewhat different impact on memory. Item analysis showed

that memory performances did strongly depend on the characteristics of single brands. These effects are currently explored to improve study materials and to control for potential confounding variables.

Presentation 2 Room 2: SenseCam as a tool for studying memory in healthy ageing Ali Mair School of Arts and Social Sciences

SenseCam is a wearable camera that automatically captures wide-angle images from the perspective of the wearer whenever a change is detected on one of the device’s inbuilt sensors (light, motion, acceleration etc.). These photographs can be uploaded to a computer and manipulated to form sequences of rich, meaningful cues for memory that are closely associated with the original autobiographical experience. SenseCam has been shown to successfully support autobiographical remembering in amnesic patients (Berry et al., 2007; Loveday & Conway, 2011) and healthy young adults (Finley et al., 2011; Sellen et al., 2007), although little is known about the determinants of its success. This project examines everyday autobiographical memory (AM) in healthy ageing using a novel paradigm that allows comparison of naturalistic remembering in young and older adults, with simultaneous exploration of some of the factors affecting event-memorability. Twenty young and 20 older adults each used SenseCam to record 12 events of their choosing and their event-memory was tested under three conditions: baseline (no pictures), forward temporal order support (all event pictures in forward order) and random temporal order support (all event-pictures in random order). Resulting memories were coded for proportions of episodic and semantic details, which were further analysed in relation to specific event characteristics. Results will be discussed in terms of their contribution to the existing literature on AM in healthy ageing.

Presentation 3 Room 2: The use of eye-tracking methodology in the research of memory in Autism Spectrum Disorders Melanie Ring School of Arts and Social Sciences

The aim of the research is to test theories about memory in ASD. Tasks address different kinds of memory like implicit memory which underlies behaviour and can not explicitly be retrieved versus explicit memory which can be consciously retrieved. Eye-tracking is sometimes considered as a measure of implicit memory. Also with eyetracking one can measure which elements participants pay attention to during the study and the test phase to see if differences in attention might be related to later differences in memory. By gathering data on where participants look at on the screen in addition to their responses we are hoping to find out at which processes difficulties in memory in individuals with an Autism Spectrum Disorder (ASD) arise. We are hoping to see if individuals with ASD show difficulties in studying or retrieving the material from memory or if they might have difficulties with both processes. In addition eye movements have the advantage of being relatively cheap and easy to measure in comparison to neuroimaging technologies. However, compared to these technologies they offer less conclusive information and have some limitations. A few of the limitations that will be discussed are: 1) What does eye-movement information really tell us? 2) Definition of fixations and its implications on the data. 3) “Cleaning” of the data 4) Problem of peripheral vision and how to deal with this in tracking eye-movements. 5) Difficulties in calibration 6) Use of eye-tracking in an abnormal population and its difficulties. Despite all these problems and limitations it will be discussed why tracking eye-movements is a useful technology and should be used when guided by hypotheses.

Paper Session 2 Presentation 1 Room 1: Relational poems – a means to highlight the polyphony within narrative; exploring influence on decision making in young women’s pregnancy stories Sarah Bekaert School of Health Sciences

Research aim: young motherhood is problematized in many contemporary western democracies. ‘Adolescence’ is constructed as a time of growing independence and preparation for the workforce. Starting a family during this time is seen to hinder these ‘tasks’, and gives rise to concerns over poverty and welfare dependency. Young people who find themselves unexpectedly pregnant tend to either abort and contracept as having a child would hinder their academic and career plans or carry on perceiving no personal loss and frequently positive gain through becoming a parent as a teenager. This research explores the pregnancy decision-making context for young women who abort a first pregnancy and continue a second as a teenager, ostensibly confounding these two pathways. Methodology: I am informed by a critical feminism that considers the role of women in a broadly androcentric society; the undervaluing of motherhood in current society, and specifically certain forms of marginalised motherhood. I also draw on ideas from Foucault to better understand the power relationships evident in the women’s narratives. Method: I explore stories of 8 women aged 17- 19, (6 Black British, one Black African and one mixed race). I use the Listening Guide, a method developed by feminist thinkers, for exploring narratives that focuses on the narrator’s voice and her relation to others and wider discourses. The Listening Guide suggests a series of 4 readings: describing plot and acknowledging reader response, tracing the ‘voice of I’ (or creating ‘I poems’), relationships, cultural contexts and social structures. This gives insight and enables interpretation of the relational context of these young women’s lives. The presentation will explore how ‘I poems’ can be used to analyse narratives. Through examining ‘I’ statements the listener maintains focus on the narrator, highlighting a sense of self with her story. I extend this process by drawing out statements for other protagonists – creating a series of poems for each narrator, enabling focus on each protagonist and consideration in relation to the main voice of ‘I’. This layered process contributes to a sense of holism and uniqueness for each participant’s story before reading across the narratives for commonalities and difference. Strengths and limitations of the method will be discussed.

Presentation 2 Room 1: Measuring cooperation and food sustainability Raquel Ajates School of Arts and Social Sciences

This paper presents the approaches taken in a cross-country study of agricultural cooperatives that explores their members’ understanding of food sustainability and where this fits in the theorisation of ecological public health. The objective of this research is to assess whether the degree of integration of consumers or retailers into farming cooperatives affects the producers’ understanding of food sustainability. Key questions to be explored include: Do members see sustainability primarily about lowering environmental impacts and costs? Or do they go beyond those green and financial aspects into other dimensions such as health and social justice or tackling the interface of current food production? Since the literature review has highlighted the limitations of financial indicators to understand cooperatives and to provide a comprehensive measure of their effects in their local communities and the wider food system, a combination of both quantitative but mainly qualitative data will be collected for this research. With a case study approach using purposive sampling, the following methods of data collection are planned: semi-structured interviews, document analysis and (a one-question) questionnaire. The choice of mixed methods also aims to overcome the barriers presented by a lack of an agreed definition or method to quantify food sustainability.

Presentation 3 Room 1: Methodological steps in developing an internet-based anxiety treatment: Current status and future plans Miriam Thiel School of Health Sciences

High prevalence rates of anxiety disorders during the postnatal period urge the need for innovation in treatment approaches. Therefore, this research project aims to adapt and evaluate internet-based Cognitive-Behaviour Therapy (ibCBT) for women with perinatal mental health issues with a focus on anxiety-related issues. This study aims to significantly contribute towards the development of available, appropriate, and accessible interventions for women experiencing anxiety in the postpartum period, and thereby potentially contribute to improved maternal and child health. The project has three primary objectives. First, it aims to develop a framework for the adaptation of ibCBT for women with perinatal anxiety to improve its suitability and accessibility. Secondly, cooperate with other research groups to develop the actual ibCBT and lastly, it aims to evaluate the acceptability of the adapted ibCBT. The adaptation and development of the ibCBT is based on recommendations by Barrera & Castro (2006) and Ritterband et al. (2003). Barrera & Castro (2006) have suggested that when developing adaptations, the following four steps should be followed: Information gathering; preliminary adaptation design; preliminary adaptation tests; and adaptation refinement. Ritterband et al. (2003) outlined nine essential steps in developing internet interventions. Based on these guidelines, a mixed method research design is in use with a major focus on qualitative methods. Planned methods included a systematic review of existing publically available ibCBT programs for anxiety, as well as focus group discussions and interviews with women suffering from perinatal mental health issues and health care professionals for both, the development of the treatment and testing its acceptability. The methods used for each step of the development will be presented and discussed in more detail along with the current status their implementing.

Presentation 1 Room 2: Rhythmic abilities of children with language impariments Ioanna Georgiadou School of Health Sciences

Children with language impairments (LI) exhibit problems with language comprehension, language production, or both. Recent studies (Brookman, A., McDonald, S., Mcdonald, D., & Bishop, DVM., 2013; Corriveau, K.H., & Goswami, U., 2009) have found that children with LI are impaired in tapping to a metronome beat, in imitating hand positions, and in moving their fingers quickly on a speeded peg-moving task. However, no studies have looked at motor skills more generally or have examined musical or speech rhythms tasks. This study attempts to fill a gap in the literature by conducting a comprehensive evaluation of the children’s motor, speech, and musical rhythm skills as they compare to the skills of typically developing children. Participants were thirty four children between the ages of 7 and 11 years of age. Children were divided in two groups: Fourteen children with LI and fourteen typically developing children (TD) who were chronologically matched to the children with LI. Children’s language skills (Clinical Evaluation of Language Fundamentals – 4th edition – Semel, Wiig, & Secord, 2006), and non-verbal intelligence (Raven’s Progressive Matrices – Raven, Raven, & Court, 2003) were tested alongside their rhythm production skills. The rhythm tasks are the most comprehensive evaluation of children’s rhythmic abilities and examine children’s rhythmic copying, periodicity, and entrainment skills. Some of the key issues to be addressed during the presentation are: 1. Research questions, hypothesis, and significance of the study 2. Methodology (participants and tasks – including audio and video examples) – advantages and limitations 3. Part of results from data analysis

Presentation 2 Room 2: The role of semantically rich gestures in aphasic conversion Judith Kistner School of Health Sciences

Background: Gestures play an important role in everyday communication (Kendon, 1997). They provide additional information to conversation partners about the meaning of verbal utterances and help to clarify even abstract concepts (Krauss & Hadar, 1999). There is evidence that gestures are not simply produced for the benefit of the listener but also support the speaker, e.g. by facilitating lexical access (Krauss, Chen, & Chawla, 1996). People with aphasia (PWA) have speech and language impairments, usually as a result of stroke. Given the role of gesture in communication, it is important to know how PWA use gesture as both an accompaniment to speech and as a compensatory modality. Several studies have investigated the use of gestures in aphasia (Rose, 2006). But only few studies have examined gesture usage in conversation (Goldin-Meadow, 2003). Aims: This study examines the natural conversational use of gestures in aphasic speech. Amongst others it addresses the following research questions: (1) To what extend do PWA employ semantically rich gestures? What impact does their semantic competence have on gesture production? (2) Do semantically rich gestures take different roles during conversation (i.e. facilitative, communicative, augmentative or compensatory)? (3) Do different topics (i.e. narrative and procedural) elicit different gesture patterns? Methods and Procedures: Twenty PWA will be recruited. Conversation samples of ten minutes will be collected based on four different topics. Video samples will be transcribed and analysed for both gesture and speech production. Semantically rich gestures (iconic, metaphoric, pantomime and emblems) and their roles will be contrasted with deictic and beat gestures. This will indicate whether participants are using gesture mainly to supplement or to replace speech. Results and Analysis: The data analysis is on-going and preliminary results will be available for presentation at the symposium. Conclusions: PWA are expected to produce a high proportion of semantically rich gestures overall to facilitate, supplement and/or replace their speech. However those who have deficits in non-verbal semantics are expected to use fewer semantically rich gestures than those with intact lexical semantics. The exact distribution of the different roles gestures can take is expected to depend on several factors. One of these factors is the number of word finding difficulties experienced by the speaker and whether these are resolved. Another one is the topic of the conversation. It is hypothesised that procedural topics are likely to elicit more iconic, pantomime and deictic gestures than narrative topics. References Goldin-Meadow, S. (2003). Hearing gesture: How our hands help us think. Cambridge, MA: Harvard University Press. Kendon, A. (1997). Gesture. Annual Review of Anthropology, 26, 109-128. Krauss, R. M., Chen, Y., & Chawla, P. (1996). Nonverbal behavior and nonverbal communication: What do conversational hand gestures tell us? In M. Zanna (Ed.), Advances in experimental social psychology (pp. 389-450). Diego, CA: Academic Press. Krauss, R. M., & Hadar, U. (1999). The role of speech-related arm/hand gestures in word retrieval. In L. Messing, & R. Campbell (Eds.), Gesture, speech, and sign (pp. 93-116). Oxford: Oxford University Press. Rose, M. L. (2006). The utility of arm and hand gestures in the treatment of aphasia. Advances in Speech-Language Pathology, 8(2), 92-109.

Presentation 3 Room 2: Let’s move: Neurocognitive mechanisms underlying sensorimotor memory Alejandro Galvez-Pol School of Arts and Social Sciences

One of the most exceptional human capabilities is the flexibility to transform our thoughts into body movements. However, the neurocognition that anchorages others’ movement is still not well understood. This study is focused on the encoding, consolidation of retrieval of sensorimotor memories for human actions. We adapted a widely used visual working memory paradigm (Vogel, 2004), recording brain activity (EEG) for visual and sensorimotor cortices, and somatosensory event potentials (SEPs) while participants performed a motor memory task. Participants were instructed to keep their gaze on a centred fixation cross, and to recall differences between a pair of arrays depicting different number of hands’ postures or polygons’ shapes. First, a cue indicated the side of the screen to be later recalled. Next, sample arrays were briefly displayed in each side and followed by a blank delay (retention interval). Lastly, a test array with dis/similar composition was presented. Here, participants were asked to recall differences (i.e., equal or different?) between the sample and test array for the cued side of the screen. Taking advantage of the contralateral disposition of the sensorimotor and visual cortex, we constructed a brain waveform that isolates the somatosensory brain activation during the retention interval, removing activity due to mere visual processing during the storage of biological stimuli vs. non-biological. The subsequent neural correlates between brain activation and accuracy for arrays with increasing cognitive demands (i.e., increasing number of stimuli displayed) are examined. EEG analyses of related paradigms offered a reliable index of memory capacity for non-biological stimuli. Besides, previous studies revealed certain neural synchrony when human postures that were embodied in the participants’ motor repertoire were observed (Calvo-Merino et al., 2010). This study integrates previous research offering a novel paradigm to examine specific mechanism for human memory, and to further explore the embodiment of others’ action.

Paper Session 3 Presentation 1 Room 1: Cooperative Control in Networks Ivana Tomic School of Engineering and Mathematical Sciences

Distributed coordination of networks of dynamic agents can be found in a broad spectrum of real life applications, such as air traffic control, satellite clusters, swarm of robots, UAV formations, etc. Cooperation between these dynamic agents plays a crucial role, so it is desired to develop an understanding of their behaviour. This paper presents a survey of recent research in consensus control of multi-agent systems, one of the most fundamental distributed technologies in the area of cooperative control. Furthermore, it describes the theoretical framework of two control schemes appropriate for a use in distributed environment. The first control design is closely related to the flocking phenomenon that can be observed in nature (e.g. behaviour of birds, fish, penguins, etc.). It is based on use of attractive/repulsive and alignment forces that generate stable flocking motion in multi-agent systems by implementing control strategies derived from carefully designed “Lyapunov” functions, corresponding to artificial potential fields. As an alternative procedure, a distributed linear quadratic regulator (LQR) control design is proposed, minimising an appropriate quadratic performance index. A centralized controller with full knowledge of the system’s state is presented and illustrated via a satelliteformation model as a motivating example; subsequently extended to a fully decentralized and distributed LQR control scheme. In this way, an optimal stabilizing controller for the overall system is obtained through the local interactions (information exchange and control) of individual agents. The paper concludes with an outline of the main research issues arising from this study which will be fully investigated in future work.

Presentation 2 Room 1: Prediction of Turbulent Diffusion Flames Using Tabulated Chemistry Kai Zhang School of Engineering and Mathematical Sciences

The current numerical study assesses the capability of employing tabulated chemistry (Imported through Ansys. Fluent ) to predict turbulent diffusion flames with varied turbulence models. The reacting jet is studied using tabulated chemistry based on equilibrium and strained steady laminar flamelets and comparisons are made with Sandia’s experimental results. The steady strained laminar flamelets obtained from opposed impinging fuel/air jets are a function of mixture fraction and strain rate (dissipation) for any pressure and initial temperatures. In order to assess the effect of strain rate (scalar dissipation), several flamelets covering the full dissipation rates and half dissipation rates are tabulated in the predictions. The predicted results are compared with experimental results in several axial locations. Result shows that there is a reasonable agreement between predicted result and experimental result. However, result shows larger discrepancy near nozzle exit which could be caused by inaccurate inlet conditions or two equation models anomaly in describing round jet flow.

Presentation 3 Room 1: Investigation of thermal effects on screw compressor clearances David Buckney School of Engineering and Mathematical Sciences

Key issues to be addressed are:  Improving realism of positive displacement compressor model  Assessing accuracy of updated model against test results Efficient, energy saving design of twin-screw type compressors, used in industrial compressor applications, relies on the ability to simulate the expected performance before manufacturing and testing. Different simulation models will have different assumptions and levels of realism; more complex models can be expensive in terms of time and resources. The objective of this work is to improve the realism and accuracy of existing fast, efficient simulation models.

Quasi one-dimensional chamber models are the industrial standard for selection and performance evaluation of screw compressors. These models generally assume rigid compressor elements when evaluating leakage areas. In this paper, a novel procedure is presented to estimate local clearance distortions by applying boundary conditions derived from a quasi 1-dimensional chamber model. Time varying boundary conditions from the nondimensional model are mapped onto 3-dimensional rotor and casing surfaces. The fluid boundary temperatures are time-averaged then used to estimate the local rotor and casing temperatures. Heat transfer assumptions that represent the extreme case for component temperature distributions are presented. The relative local change in clearances between rotors and casing can then be estimated without the need for finite element analysis (FEA). Revised leakage areas which account for thermal distortions are fed back into the chamber model to provide improved performance predictions. The accuracy of the corrected predictions is assessed using test results from a twin-screw compressor.

Presentation 1 Room 2: A Service Selection Framework for Security-Aware Web Service Composition Ahahedeh A.Khani School of Informatics

Abstract—Service Oriented Architectures (SOA) allow the development of software systems by the composition of independent web-services, which results in less development effort, time consumption and cost. These webservices offer their available operations through an interface. For example, a bank might expose a credit card processing service allowing businesses to charge their customer’s credit cards for the service they offer. This allows companies and organizations to only implement their core business and outsource other service components over the Internet either pre-selected or on-the-fly. However, if the outsourced web-services suffer from security vulnerabilities they will be at risk of security attacks, which can in turn cause unavailability, loss of confidentiality and/or integrity of the composite service. Therefore, selecting a suitable service is one of the major challenges in these architectures. Service selection, based on the Quality of Service (QoS) criteria, has received great attention from researchers. However, less research has been done on security-based service selection. Web-services with security vulnerabilities will be at risk of security attacks. Service selection frameworks should consider both QoS and security level of the available services. Our research focus is on increasing the security of composite web-services through a broker-based architecture and it is in early stages of design and development. To achieve our goal we are going to combine various methods, each tackling a specific issue in the area of software engineering. This framework selects diverse sub-services that offer the required functionality, while being informed of their current security vulnerabilities identified through penetration testing, then combines them to achieve a more secure composite service.

Presentation 2 Room 2: Double Reading of Mammograms: Improving Performance by Effectively Pairing Readers and Identifying and Discussing Difficult Cases Marwa Gadala School of Informatics

Purpose: Double reading, interpretation of a medical image by two pairs of eyes, is standard practice in breast cancer screening programs in over 12 countries. We retrospectively investigated whether its benefits could be increased by (1) forming complementary reader pairs and (2) isolating difficult cases for discussion. Method: We used data from a UK clinical trial where 50 readers each read 180 mammograms - 60 with cancer and 120 normal. First, we formed complementary reader pairs(HL) by grouping high sensitivity readers (H) with low sensitivity readers (L), and simulating their performance using the OR recall rule. Their performance was compared to those of homogeneous(LL+HH) and random(HH+LL+2LH) pairs. Second, we used the 4-point scale to identify unsure cases. Rather than double read all cases, unsure cases were discussed and sure cases were double read. Discussion was ideally simulated as yielding the correct decision, and then more realistically according to authority or confidence. Statistical significance was determined using Welch’s t-test, and 95% confidence intervals (reported in square brackets).

Results: Complementary pairing significantly increased sensitivity by 4.4% (p=9.85e-9), compared to homogeneous pairing, with no significant effect on specificity. Unsure cases (16%) were nearly five times more likely to be incorrectly diagnosed. Double reading all cases with an average reader yielded: sensitivity 0.834 [0.822, 0.847], specificity 0.719 [0.698, 0.739]. In comparison, ideal discussion of unsure cases with the best reader yielded: sensitivity 0.857 [0.844, 0.869], specificity 0.829 [0.807, 0.850]. Modelling discussion with an average reader or according to realistic factors sometimes increased sensitivity, and always increased specificity with varying significance and non-significance. Conclusion: Pairing readers simply according to convenience could be significantly less effective than pairing according to ability. A diverse rating scale can identify uncertain cases for discussion, and thus reduce unnecessary biopsies by as much as 11%. Our simulations suggest these protocols are worth experimental exploration.

Presentation 3 Room 2: Influence of colour and luminance on visual working memory – a study using EEG Maciej Kosilo School of Arts and Social Sciences

To perform everyday tasks people rely on working memory (WM) to keep track of what they are doing and what they need to do next. Here, we are interested in visual WM. There are three stages to WM: encoding, when the memory representation is created, maintenance, when we hold the information in mind, and retrieval of that information later on. This experiment aims to understand how the encoding stage of WM can affect subsequent performance. It has been suggested that everyday vision benefits from luminance inputs (Bar, 2003). However, it has been not investigated so far whether such low-level visual properties can affect the way memory representations are created. Our project aims to understand how basic visual information, like colour and luminance, affects WM encoding and whether this is related to a successful performance in those tasks. Our task requires participants to remember simple shapes. To introduce low–level differences between the stimuli, shapes are defined along different directions in colour space. The purpose of this is to bias the brain’s processing of the stimuli, so that it uses different kinds of information to form a mental representation. For example, we could present shapes that are defined predominantly by changes in luminance, or, alternatively, by changes in chromaticity (colour).We aim to explore whether those conditions will have a differential effect on participant’s performance in terms of accuracy and reaction times. Additionally, we aim to compare whether neural activity was modulated by our manipulations as well. Because encoding into working memory happens almost instantaneously, we will utilise electroencephalography (EEG) to achieve this goal. EEG is a technique that allows to record electrical activity of the brain from the surface of the head as it happens, in real time. As the data collection is in progress, some preliminary data will be presented and future directions discussed.

Paper Session 4 Presentation 1 Room 1: Children’s Language skills: Are we sure this is what we are talking about? Hannah Hockey School of Health Sciences

The paper will begin by presenting the study’s aims and rationale, to give a context for the main discussion. Aims of the study To investigate the skills involved in pre-school children’s ability to repeat made-up words, real words and sentences. Specifically, children’s performance in the following areas: • Knowledge of the words • Pronunciation • Verbal short-term memory Rationale for the study Repetition tasks are often used in speech and language therapy research to pick out children who have language disorders. Through identifying the skills influencing repetition performance, speech therapists would be better able to plan targeted therapy for the children’s difficulties. The paper will then describe the tasks used to assess the language skills of the children (aged 3 and 4 years). These include: • A picture-naming task • A picture-pointing task • A repetition task • A verbal short-term memory ‘span’ task • Existing standardised language assessment (Pre-School Language Scales-4) and ‘non-verbal’ assessment (British Ability Scales II) The paper will discuss some of the confounding methodological variables that interfere with collection and interpretation of the data: • challenges in collecting quantitative data from young participants and attempts made to overcome these challenges • strengths and limitations of the tasks chosen for the study • strengths and limitations of the methods used to score the data The paper will question whether a test of ‘performance’ serves as an indication of an individual’s language ’ability’. It will discuss briefly the wider implications (research and clinical) of this. Finally it will discuss other methods that may be available to researchers and therapists when assessing language skills in young children.

Presentation 2 Room 1: What is ‘normal’ in discourse production – and can we measure it? Madeleine Pritchard School of Health Sciences

Background. Discourse describes language beyond a single simple sentence that speakers use for a specific purpose (Armstrong, 2000). Discourse is an everyday use of language: telling a friend about the weekend, giving instructions to complete a procedure, or describing a scene are all common communicative activities, requiring complex high level skills. People with aphasia, a language difficulty occurring following stroke, often have difficulties with discourse, which impacts on their daily life (Wright, 2011). Despite this, assessment and treatments focusing on discourse are not widely practised (Rose, Ferguson, Power, Togher, & Worrall, 2013). This may be due to a poor understanding of how speakers use language in discourse, and which factors contribute toward ‘good discourse’.

Challenges. A range of methods are available which describe how speakers convey meaning in discourse. For example, macro measures look at the information level of a discourse, and describe aspects such as how much information a speaker presents; how well each utterance relates to the overall topic; and the amount of detail a speaker gives. Each of these may be useful in identifying what makes ‘good discourse’, and provide a useful way to measure a speakers’ impairment, or the outcome of treatment. However, there are a number of challenges relating to these analyses and their use. Firstly, these analyses are not explicitly linked to models of language or cognition. This means it is unclear what they indicate, and how they could be used to measure a speakers’ difficulties or progress. Secondly, there is variation when the analyses are used, making comparison of findings challenging. This paper will present a critique of these analyses, and aims to 1) Explain the analyses which compare the macro level of discourse, and how these relate to models of language and cognition 2) Compare the relative strengths and weaknesses of each analysis Identify the practical challenges of implementing these analyses, for the purposes of research and clinical practice.

Presentation 1 Room 2: Moral Dynamics in Everyday Life: How morality evolves in time? Albert Barque-Duran School of Arts and Social Sciences

How individuals face the ethical uncertainty of their lives? Donating to charities, volunteering, recycling, buying fair trade products, donating blood… People are confronted with a conflict between cultivating the moral selfimage and pursuing self-interest. How this conflict is resolved? In this study we want to replicate some recent and interesting results in the field such as Moral Balancing (behaving ethically or unethically decreases the likelihood of engaging in the same type of behavior again later) and Moral Consistency (engaging in an ethical or unethical behavior increases the likelihood of engaging in the same type of behavior later on). Furthermore, our aim is to provide new empirical data in the field. In concrete, the purpose is to explore if these moral patterns are maintained over time. To do so, we are going to ask participants on their frequency of engaging in a promoral behaviour (7-point scale: -3=very unlikely, +3=very likely) through a sequence of moral scenarios that are going to be presented. In Study1 we manipulate the mind-set of the participants (Outcome-Based, whether an act is morally right depends on the consequences of that act vs. Rule-Based, what makes an act right is its conformity to a moral norm), as it seems to moderate their ethical behaviours. In Study2 we manipulate their mind-set and their memory (last moral recall of an action) before starting the experiment, as memory seems to moderate ethical behaviours too. In Study3 we manipulate their mind-set and their memory on each decision they take. We have a control group with no manipulation at all. The results should tell us if moral patterns such as balancing-consistency are maintained over a sequence of moral scenarios. In other words, how moral behaviours evolve in time. Research methods and tools used will be discussed. One the one hand we will present the use of digital platforms (Amazon Mechanical Turk) as a way to recruit and test participants. On the second hand we will present the idea of using computational models based on quantum probabilities to show patterns and make predictions in the field of cognitive sciences.

Presentation 2 Room 2: Using Interpretative Phenomenological Analysis to explore the lived experiences of HIV positive mothers in Kingston, Jamaica Gayle Clifford School of Health Sciences

My research uses Interpretative Phenomenological Analysis (IPA) to explore Jamaican HIV positive mothers’ experiences of talking to their seronegative children about their HIV status. IPA is a relatively new approach to analysing individual experience developed by Jonathan Smith in 1996, and is influenced by phenomenology, hermeneutics and idiography. To date there is scant evidence on the experience of maternal ‘disclosure’ in resource-constrained contexts. In this presentation I will discuss the benefits and limitations of IPA both generally, and in relation to my particular study. IPA allows us to understand the experiences of particular groups (‘gives voice’); to develop and evaluate services and interventions; and to situate and understand people’s experiences in their socio-cultural contexts (Smith, Flowers & Larkin 2009). IPA provides detailed information on the experiences of often under-represented groups, whilst recognising the impact of the researcher, the ‘double hermeneutic’ (Smith & Osborn 2003). I conducted interviews in Kingston, Jamaica with 15 HIV positive women recruited through an NGO and an HIV clinic, as well as interviews with 12 HIV professionals. Existing support is limited and tends to focus on disclosure to partners and seropositive children. Preliminary findings show that women are concerned about and unsure of how to manage communication about HIV with their children, including seronegative children. Women discussed their desire to be open about their status alongside their fears and experiences of discrimination, ostracism, loss of livelihood, relationships and opportunities. IPA is well suited to my study as it seeks to examine the personal experiences of a particular group. As a stilldeveloping approach, IPA allows for flexibility and creativity on the part of the researcher (Willig 2001). The phenomenological and hermeneutic aspects of IPA recognise the importance of participant and researcher interpretation as well as the cross-cultural issues of conducting research as an ‘outsider’ in Jamaica.

Posters 1. Chemotherapy induced cognitive changes in colorectal cancer patients Marie-Rose Dwek School of Health Sciences

Chemotherapy drugs can cause severe side effects, with many cancer patients reporting a condition anecdotally known as “chemofog” or “chemobrain”. Research suggests that chemotherapy causes decline in cognitive functions including memory and attention which could affect patients’ quality of life and daily functioning [1-7]. However, the natural course of any cognitive decline and whether or not it causes observable difficulties for patients over time is relatively unknown. The majority of research studies have explored cognitive function in cancer patients after treatment has been completed, producing inconclusive results. Few studies have measured patients’ cognitive function before chemotherapy [8, 9]. Measuring cognitive function both before and after chemotherapy would identify any changes occurring during treatment. There is also a lack of comparison groups in existing studies (e.g. nonchemotherapy cancer patients) against which to compare cognitive function scores. Despite reports from mixed-gender cancer populations that they experience the same problems, the majority of research has focussed on breast cancer patients, precluding an exploration of gender differences in cognitive decline. Colorectal cancer (CRC) is prevalent in many developed countries, affecting men and women almost equally [10]. With chemotherapy administered as standard, CRC patients have a comparatively high survival rate, which is ideal for a longitudinal study looking into chemotherapy-induced cognitive changes. This study aims to identify the extent and impact of chemotherapy-induced cognitive decline and the supportive care needs of the CRC population. A longitudinal comparative design will be used to collect data using questionnaires and neuropsychological assessments as recommended by the International Cognition and Cancer Task Force [7] including the Hopkins Verbal Learning Test-Revised; Trail Making Test and Controlled Oral Word Association pre-, mid- and post- chemotherapy. A total of 146 CRC patients from participating hospitals will be assessed; half requiring chemotherapy and half non-chemotherapy surgical patients. References 1. Vardy J., Wefel J.S., Ahles T., Tannock I.F., Schagen S.B. Cancer and cancer-therapy related cognitive dysfunction: an international perspective from the Venice cognitive workshop. Ann Oncol. 2008; 19(4):623-9 2. Schagen S.B., van Dam F.S., Muller M.J. et al. Cognitive deficits after postoperative adjuvant chemotherapy for breast carcinoma. Cancer 1999; 85: 640–650. 3. Brezden C.B., Phillips K.A., Abdolell M. et al. Cognitive function in breast cancer patients receiving adjuvant chemotherapy. J Clin Oncol. 2000; 18: 2695–2701. 4. Ahles T.A., Saykin A.J., Furstenberg C.T et al. Neuropsychologic impact of standard-dose systemic chemotherapy in longterm survivors of breast cancer and lymphoma. J Clin Oncol. 2002; 20: 485–493. 5. Wieneke M.H., Dienst E.R. Neuropsychological assessment of cognitive functioning following chemotherapy for breast cancer. Psychooncology 1995; 4: 61–66. 6. van Dam F.S., Schagen S.B., Muller M.J et al. Impairment of cognitive function in women receiving adjuvant treatment for high-risk breast cancer: high-dose versus standard-dose chemotherapy. J Natl Cancer Inst 1998; 90: 210–218. 7. Wefel J.S, Vardy J., Ahles T., Schagen S.B. International Cognition and Cancer Task Force recommendations to harmonise studies of cognitive function in patients with cancer. www.thelancet.com/oncology. Published online February 25, 2011. DOI: 10.1016/S1470-2045 (10) 70294-1 8. Falleti M.G., Sanfilippo A., Maruff P., Weih L., Phillips K.A. The nature and severity of cognitive impairment associated with adjuvant chemo- therapy in women with breast cancer: a meta-analysis of the current literature. Brain Cognition. 2005; 59(1): 60–70. 9. Anderson-Hanley C., Sherman M.L., Riggs R., Agocha V.B., Compas B.E. Neuropsychological effects of treatments for adults with cancer: a meta-analysis and review of the literature. J Int Neuropsychol Soc. 2003; 9(7): 967–982 10. Arndt V., Merx H., Stegmaier C., Ziegler H., Brenner H. Quality of life in patients with colorectal cancer 1 year after diagnosis compared with the general population: a population-based study. Journal of Clinical Oncology. 2004; 22(23):482936.

2. The Impact of Health Insurance Coverage on Health Care Services among Patients with Chronic Diseases in the Eastern Province of Saudi Arabia Afnan AlJaffary School of Health Sciences

Abstract Generally, health insurance has been recognised as a necessity worldwide but it has only recently been considered a major priority by Arab states. Saudi government hospitals are not properly equipped to serve a rapidly growing and ageing population; nor are they ready for the changing patterns of chronic conditions. Decision-makers in Saudi Arabia are planning to restructure the healthcare system and introduce Council for Cooperative Health Insurance (CCHI) scheme for Saudi citizens with a view to improve the quality of healthcare (Alnaif 2006). The first stage of this scheme which is providing Saudi and non-Saudi citizens who are working in the private sector with health insurance coverage is already applied. However, the second and third stages of CCHI scheme which aim to cover all citizens and pilgrims in Saudi Arabia are still pending (Al-shraqi & Abdullah, 2012). In order to inform Saudi healthcare system policy makers around the impact of of CCHI implementation on healthcare services, it is necessary to conduct further research in the area (Berhie 1991). To date research on health insurance in the region is relatively scarce compared with that available in the developed countries. This study is a mixed approach study which mainly based on patient-centred approach; in the hope those patients’ health care needs will be addressed and met by the provision of optimal healthcare services possible for the entire Saudi community. The study aims to measure the impact of health insurance coverage on patients’ health seeking behaviour and the viability to introduce it in Saudi health market that is mainly served by public health insurance in the eastern province of Saudi Arabia. The research will be based on a construction of a questionnaire survey and semi-structured interview, which will be distributed to around 400 Saudi households in the eastern province of Saudi Arabia. References: Almalki, M, Fitzgerald, G & Clark, M 2011, ‘Healthcare system in Saudi Arabia: an overview’, Eastern Mediterranean Health Journal, vol. 17, no. 10, pp. 784–793. AL-Sharqi OZ,Abdullah MT 2012, “Diagnosing “ Saudi Heath reforms: is NHIS the right prescription?” , International journal of Health Planning and Management. DIO:10.1002/hpm.2148. Berhie , G 1991, Emerging Issues in Health Planning in Saudi Arabia: The Effects of Organisation and Development on The Health Care System, Social Science Medical Journal, 33,no.7. pp. 815-824

3. Pre-verbal and early verbal communication in siblings of children with language impairment or Autism Spectrum Disorder Helen Cain School of Health Sciences

Background The research literature on Autism Spectrum Disorders suggests several early markers that may be present in infants who later go on to develop ASD, including deficits in eye contact, imitation, gesture, joint attention (shared focus with an adult on the same object), requesting behaviours and play skills. By contrast, although clinical markers for Specific Language Impairment have been identified in school age children (including nonword repetition and tense marking abilities e.g. use of the past tense –ed when describing past events), it is unclear whether there may also be earlier markers present in the early- and pre-verbal communication of infants who will later be diagnosed with language impairments. There is also a debate in the literature about whether ASD and SLI represent overlapping diagnostic categories, which could be informed by work comparing the early developmental trajectories of the two disorders using the same set of measures. Methodology The study is using a prospective longitudinal design to begin investigating language impairment and Autism Spectrum Disorder before diagnosis. An at-risk sibling cohort will be recruited, as the recurrence rate of language impairment and ASD among siblings is higher than in the general population.

The study will recruit 3 groups of infants: a language-high-risk group, an ASD-high-risk group, and a control group of younger siblings of children with typical development. These infants will be assessed for the first time at around 8 to 14 months of age, and will then be reassessed a year later when they are about 24 months of age. The assessment protocol for the first time point will include video recording of the infant’s spontaneous communication with their caregiver during play, parent report measures of vocabulary and non-verbal communication, and a novel dynamic assessment of non-verbal and early verbal communication skills, including imitation, gesture, joint attention and receptive language. Results Preliminary results will be presented based on analysis of the Time 1 assessment data.

4. The role of image editors in the image selection process in online journalism Sylwia Frankowska-Takhari School of Informatics

Background information Most research into image users’ needs and searching behaviour in the context of journalism focuses on image needs of news professionals. From literature, it is known that when selecting images for news articles, journalists and image editors take into consideration the content of images, as well as a variety of other visual and non-visual image attributes, and the final selection is influenced by editorial guidelines (guarding the journalistic ethics), and also frequently by personal preferences, or ‘environmental’ factors such as: image cost, time on task. Research problem Little is known about image needs of news readers, and whether in the image selection process professionals take these needs into account, or whether they compromise them with e.g. ‘environmental’ factors and choose images that fill the space on the page rather than meet readers’ needs. Uncovering readers’ image needs will contribute to creating a complete model of the image selection process, and inform the development of image retrieval systems to enable news professionals to select images that meet readers’ needs and optimise other factors. Research design To investigate image selection criteria used by professionals, and consumers’ needs, the researcher has designed a 3-phase study: Phase 1. In-depth interviews and observations with image editors at their work places. Phase 2. A task-based study with regular news readers who will be engaged into image selection tasks simulating work tasks of image editors. Phase 3. A task-based study with regular news readers to evaluate image choices made by image editors (in Phase 1.), and the non-professionals (in Phase 2.). The researcher is now in Phase One of the study. 12 news professionals filled in an entry questionnaire and participated in individual semi-structured depth interviews, 9 of the 12 were further invited to take part in observation sessions in their workplaces. In the observations, participants were encouraged to ‘think-aloud’. This allowed the researcher to get a better understanding of the actions and decisions that were being made. Additionally, during the research sessions the researcher compiled a collection of artefacts i.e. examples of images considered for selection and those selected for content in the sessions. Learnings Triangulating the research methods enabled the researcher to collect large amount of data, and cross-validate findings. Conducting the sessions in the context of participants’ workplace allowed the researcher to observe participants while they were working on real-life tasks, in their ‘natural’ environmental. However, it was also observed that the think aloud protocol significantly influenced the time required to complete each job task.

Preliminary findings from Phase 1. The preliminary data analysis indicates that the main selection criteria used by professionals are: image content, novelty, quality, and colours. However, in most cases the final selection is made to suit personal preferences of the professionals.

5. Measuring the effects of negative emotions on the production of false memories Cassandra Bland School of Arts and Social Sciences

False memories are relatively new and fascinating phenomena and there is much research into the various influencing factors. The factor of interest in this research is negative emotion. Negative emotions have been shown to increase susceptibility to false memories, and even more so when the experienced emotion is congruent to the emotional content of the remembered material. Current research however has retained a fairly general view of negative emotions, and little attention has been paid to more specific, individual, emotions. This research aims to drill deeper into negative emotions and the effect they have on false memories. Many different paradigms in which false memories can be investigated have been developed, however they each produce significantly different results. Spontaneous false memories are investigated, most commonly through use of the DRM paradigm. Lists of words, that are all highly associated to one non-presented word, are typically recollected along with the “false” memory of the non-presented word. These word lists can then be imbedded into stories, increasing ecological validity, but also increasing the vulnerability to confounding variables. Implanting memories is an alternative technique, with significant differences on the production of false memories, and many more ethical issues. With regards to the study of emotions, there are several methodological issues regarding, how emotions can be manipulated, induced, and measured, and if these procedures are ethical. With emotion induction it is often the case that multiple emotions are elicited simultaneously, making it difficult to drill down into the effects of each specific emotion. This research will combine many of the mentioned methodologies in order to give a much broader, and clearer, picture of what is going on.

6. Are these reviews trustworthy? Investigating the role of user interface elements in signalling reviews’ trustworthiness and the effect of user belief Dara Sherwani School of Informatics Centre for Human Computer Interaction Design

The increasing number of systems that provide user-generated reviews has dramatically changed the way users make purchase decisions. For example, TripAdvisor, the most popular travel review website, allows users to assess services prior to buying them, however, users run the risk of making a wrong purchase by trusting biased reviews. One way to solve this problem is to automatically detect untrustworthy reviews using models and algorithms. However, this has not been successful or reliable to date. Our work takes a different approach, inspired by Riegelsberger et al. (2005) for supporting user trust, and explores how user interface (UI) elements can signal trust-related factors and enable users in distinguishing between trustworthy and untrustworthy reviews. Our work takes users’ backgrounds into consideration in their assessment of trustworthiness, especially dispositional trust which refers to individual’s consistent tendency to depend on others (Hsu, 2008). In order to investigate these issues, we conducted an exploratory study with 16 participants in which they interacted with a computer-based prototype, similar to TripAdvisor. We manipulated review trustworthiness through degrees of review quality, reviewer expertise and reviewer bias. Participants were asked to investigate hotels for a holiday based on other travellers’ reviews. We gathered both quantitative data capturing participants’ dispositional trust and their ratings of review trustworthiness, and qualitative data through participants’ verbalisations as they were “thinking -aloud” as they read the reviews. This approach offered advantages especially because of the qualitative data that provided in-depth understanding of the way users trust reviews based on UI elements. It also revealed clear differences between participants with high and low

dispositional trust in assessing reviews’ trustworthiness. However, the effect of dispositional trust should be validated in future work since it was captured quantitatively. This would require larger scale studies. Our results suggest implications for designers of systems that provide user-generated reviews that could help users in assessing reviews’ trustworthiness. Our work also points the way to future research in better understanding the role of users’ backgrounds and appropriate trust signals in assessing review trustworthiness.

7. Measuring Makers Making – Designing an End-User Programming study of Arduino development environments Tracey Booth School of Informatics Centre for Human Computer Interaction Design

As part of the growing Maker Movement, a trend towards open source hardware has reinvigorated interest in hacking and tinkering, through low-cost electronics prototyping platforms such as Arduino, drawing in people with little engineering or programming experience - artists, researchers, hobbyists and young people - who must then learn and apply skills in both – a twin challenge. I present a methodology and tools used for a HCI study into the relative effectiveness and user experience of different programming environments – visual (graphical) and textual – for adult novice end-user programmers of Arduino. This exploratory study, a precursor to my current research in this domain, was the first to explore end-user programming (non-professional programming) in the context of physical (electronic) prototyping. A withinsubject experiment, it exposed participants to both visual and textual programming environments for the completion of set tasks. Several measures were taken, including: • • • • •

Task completion (programming efficacy) Self-efficacy ratings (general and post-task) - using a standard questionnaire (Compeau & Higgins, 1995) Perceived workload ratings (post-task) - using the NASA-TLX standard questionnaire Product Reaction Card selections (post task) – adapted from the Microsoft Desirability toolkit (Benedek & Miner, 2002) Learning Barriers – a qualitative analysis of task video transcripts, using Ko’s Learning Barriers (2004), determined types and frequencies of barriers encountered per environment.

I present the above methods/tools as used in the context of my research and outline their benefits and drawbacks, reflecting on my experiences, including implications for analysis/reporting. I also refer to some of the general issues/challenges involved in designing programming-related Human-Computer Interaction studies.

8. The use of simulators in the assessment of computational models of learning and behaviour Niklas Kokkola School of Informatics

Numerous issues have been brought up by the use of simulators in verifying the predictive power of computational theories of animal learning and behaviour. The latest special issue of Learning & Behaviour details these obstacles. Specifically the conflicting demands of designing useful and simple algorithms, which replicate empirical findings and the plausibility of the equations used in terms of agreed upon neurobiological and psychological structures and processes. Similarly, the issue of maintaining the research standard of the reproducibility of simulator results by other researchers is important in both verifying claims and fostering collaboration. This proceeds through sharing of simulator source code among researchers. The necessity of using of fixed numerical constants from simulation to simulation when assessing performance of a simulator in predicting a multitude of different phenomena is crucial to rigorous standards of theoretical verification. It is

often ignored or only partially followed, though sometimes for valid reasons, which can produce confounding factors when comparing computational models. Additionally, the establishment of consensus on critical phenomena, which need to be covered by any comprehensive model of learning and behaviour, and the difficulties inherent to this process, are detailed. Finally, the assessment of complexity of a model in contrast to performance in terms of the breadth of phenomena, which can be accounted for with the aforementioned fixed parameters, will be analysed. Recent computational models will be compared in terms of their complexity, predictive accuracy and breadth, consistency in using fixed numerical constants for explaining phenomena, neuropsychological plausibility, and accessibility. Furthermore issues of consensus in empirical research will be analysed through comparisons of reports of empirical learning and behavioural phenomena with the consistency of those results with the researcher’s own models.

9. How does the tangible object affect motor skill learning? Milena Markova School of Informatics Centre for Human Computer Interaction Design

Although tangible user interfaces have gained recent popularity in research, their impact on certain types of learning has not been studied extensively; therefore the effects of their use are not fully understood yet. This poster will discuss a classification of tangible user interfaces for learning and the design of a study which looked at how two different tangible objects in a tangible user interface affect motor skill learning. The study compared motor skill learning in two conditions, one of which featured a tangible object whose shape was manipulated in order to resemble a real life object. The hypothesis was that the participants using a tangible object more similar to a real life object would achieve higher competence in a motor skill than the participants using a tangible object less similar to a real life object. An empirical study was designed to investigate the hypothesis, which compared learning how to play baseball, and more specifically bat, using Nintendo Wii Baseball. The study featured a between-subjects design. Thirty five participants were recruited, mainly from the student body of City University London, and each participant was randomly assigned to one of two conditions – control group (using an object which expressing a verb metaphor, where only the action performed with the tangible object was similar to its digital representation), and an experimental group (using an object expressing a noun and verb metaphor, where both the action performed with the tangible object and its shape were similar to its digital representation). The participants were asked to perform three tasks, five times each, which were the same for both groups. A repeated measures method was used – participants attended three sessions, each one week apart. This poster summarizes the current state of the data analysis stage of the study, as well as plans for future work.

Please remember to complete a voting form for all the presentations you attend and for each poster you view.