Introduction: There is emphasis on increasing patients’ Physical Activity (PA) to reduce disability and promote independent living. Therefore a new computerised system based on real time location technology called the Rehabilitation Mobility Measurement System (RMMS) was developed to overcome limitations of the current activity monitoring methods and measure PA continuously and unobtrusively. The study objectives were to evaluate the psychometric properties of RMMS and to explore early stage functional recovery after stroke in a rehabilitation unit and at home. Methods: Each participant wore a radio-frequency identification tag with an in-built motion sensor on their unaffected wrist. Walking-aids and transport equipment were also fitted with tags. All areas accessed by patients were fitted with infra-red room locators. The tags transmitted movement and location signals to a computer having customised software programs for data processing. Descriptive statistics and graphs were used for analysis. Results: The RMMS was very reliable (all ICC > 0.90) and demonstrated high level of agreement on validation with observational methods. Longitudinal PA was measured successfully in the rehabilitation unit for 52 patients over 64±53 days. Outside of therapy sessions, patients spent 85% of the waking day in their own rooms undertaking limited high level activities (15%).The average mobility (walking or moving around) was 15 minutes per day only and was strongly correlated with Barthel Index and modified Rivermead Index scores on discharge (spearman’s rho=.-70, p=0.00) accounting for ≥ 43% of variation in these scores. Conclusion: RMMS was a reliable and valid tool for measuring mobility; a key factor influencing early stroke recovery. The small amount of time spent active strongly suggests that better organisation of time outside therapy sessions is warranted to maximise daily PA of in-patients. RMMS could be used for motivational feedback for patients and clinicians to ultimately enhance functional activity during rehabilitation in a stroke unit.

Application of GPS technology in elite level soccer is a growing area of research. This thesis comprises an examination of current practice in elite youth level soccer, and a critical examination of the potential applications in the PlayerLoadTM measure to quantify the biomechanical demands of match play. The thesis comprises four experimental studies that consider the development of monitoring biomechanical intensity in training and/or competitive matches. The first experimental study provides a critical examination of the biomechanical specificity of training drills relative to competitive match play. This study utilised the performance metrics as collated on a daily basis by the football club. Specifically, in relation to tri-axial accelerometry the measurement of PlayerLoad was restricted to total accumulated loading. Additional parameters related to distance and velocity parameters were also examined. ‘Small-Sided Games’ generated similar values to 90-min matches for PlayerLoad (standardised for duration) and total distance covered. However, these drills failed to provide a valid demand in terms of high-intensity running, which was most valid in ‘Movement Pattern’ drills. Drills described as ‘Possession’ and ‘Game-Related’ failed to match the mechanical demands of match play. The implications of these findings relate directly to the micro-design of the training week, and the monitoring of player performance. The correlation between PlayerLoad and distance covered was stronger in small-sided games (r=0.92) than in regulation 11 vs 11 match play (r=0.37), highlighting mechanical issues in the calculation of PlayerLoad. The smaller pitch size is likely to promote a greater frequency of speed and or directional change, and as such the summation principle applied to generate a “total” or 3- dimensional loading value is limited. 6 In the second experimental study the analysis of tri-axial accelerometry was extended to provide a uni-axial consideration of PlayerLoad. Biomechanically, this is analogous to analysing each force vector rather than the development of a “total” kinetic parameter based on a summation principle. This uni-axial analysis of mechanical loading was first applied to the influence of playing age via a comparison of the U16, U18 and U21 squads within the same club. The U16s performed the greatest total distance, primarily in the lower speed zones. Correlation between PlayerLoad and total distance ranged from r=0.26-0.56, for the three age groups, with evidence of higher coefficients in the U16 group. The U18s exhibited the greatest PlayerLoad, evident in each movement plane. Uni-axial analysis highlighted a higher contribution from medio-lateral loading in the U18s, indicative of greater lateral movement. This finding might also relate to the higher injury incidence observed in this U18 age group. The practical applications of this study relate to the transition of players through the academy structure and into senior football. The unique movement patterns identified by a uni-axial analysis of PlayerLoad highlights potential in the greater analysis of movement. This uni-axial analysis was extended in the third experimental study to further examine issues in the movement profile with a consideration of the influence of playing position on mechanical loading. Whilst not generalisable beyond this team and playing strategy, attackers covered the greatest (total and high speed) distance, whilst midfielders exhibited the greatest load across all movement planes. Correlation between PlayerLoad and total distance was position specific, forwards and midfielders recorded values of r=0.74 and r=0.16 respectively. Playing position categorising defenders, midfielders and attackers failed to identify the impact of positional width on the biomechanical demands of match play. The traditional grouping of playing units might therefore be considered in terms of individualising training programmes. The distinction between distance covered and PlayerLoad is consistent throughout the first three experimental studies, with a low correlation in part explained by the calculation used to 7 quantify PlayerLoad. In the final experimental study the PlayerLoad calculation is critically examined beyond the uni-axial nature of acceleration. Having previously examined the summation principle, the failure of the PlayerLoad calculation to consider magnitude of acceleration is examined. The instantaneous change in acceleration is not influenced by the magnitude of acceleration, and in the final study a novel iLoad parameter is introduced which is analogous to the iEMG parameter utilised widely in electromyography. This parameter considers the integral of the acceleration-time curve. Further, the sign principle is critically examined, with the PlayerLoad calculation negating all negative values and thus making all movements forward, to the right, and upward. By considering both positive and negative values the tri-axial accelerometer has the capacity to differentiate between medial and lateral movement for example, with clear implications for the monitoring of performance and injury risk. This novel biomechanical analysis was applied to an examination of fatigue during match play, which has implications for both performance and injury. Over 15min segments of match play, fatigue did not influence the anterio-posterior or medio-lateral loading but there was a significant decrease in vertical load. There was also evidence of movement asymmetry in each plane, favouring movements forward and to the left. Correlation between iLoad and total distance was r=0.19. In conclusion, the thesis evaluated PlayerLoad and critically discussed the mechanical specificity of training activities. Furthermore, use of uni-axial load highlighted differences in positional demands and the influence of age group on GPS variables. Critical evaluation of PlayerLoad calculation aimed to highlight the deficiency of tri-axial acceleration of the formula. Thus, iLoad further developed calculation to refine movement quality data to examine fatigue. By adopting principles analogous to kinetic analyses in force platform and electromyography, additional analysis parameters may be defined which provide greater depth 8 of information in movement quality. The implications in movement asymmetry also have implications for the monitoring of injury risk.

The independent effects of acute heat and hypoxic stress on human physiological function and performance are relatively well documented. Although in the field these environmental stressors rarely occur in isolation the effects of combined or sequential exposure to them has not been extensively studied in humans. Animal models have however shown that acclimation to one stressor can induce ‘cross acclimation’ a positive adaptive response upon exposure to a different stressor. The three studies within this thesis were conducted in humans to assess how exposure to acute and repeated exposures to heat affects the later physiological and cellular responses to acute exercise in normobaric hypoxia. A possible site for any cross-acclimatory affects and conferred cellular tolerance resides in the heat shock response (HSR) and the increased expression of heat shock proteins (HSPs). The 72 kilodalton HSP, HSP72 has been implicated in heat acclimation mediated cross acclimation in rodent models, and also shown to be important in the human adaptation to heat and hypoxic stressors. Study One determined the physiological and HSR to exercise in both heat (HEAT; 40°C) and hypoxia (HYP; FIO2 0.14) alone, and in combination (COM) as well as a normothermic normoxic control (NORM). 24 hours after the initial exposure a hypoxic stress test (HST; 15 minutes of seated rest and 60 minutes of cycling exercise at 50% normoxic peak) was conducted to determine what effect the prior stress exposure had on both whole body physiological responses and the cellular HSR. It was hypothesised that the stressor that elicited the greatest physiological strain and HSR on day one would have the biggest effect on reducing physiological strain in a subsequent HST. Twelve male participants completed 4 trials consisting of a 15 minute rest period in normoxic temperate conditions, followed by 30 minutes seated rest and 90 minutes cycling exercise at 50% Npeak within NORM, HEAT, HYP and COM. 24 hours after completing this exercise bout, participants undertook a HST. Exercise duration was reduced in HEAT (78 ± 12mins), HYP (81 ± 13mins) and the CON (73 ± 19mins) trial compared to the NORM (89 ± 3mins). HR and core body temperature (Tcore), and thus physiological strain, were greater in the HEAT and COM trial compared to HYP alone. This response was also observed with post exercise monocyte HSP72 (mHSP72). Basal HSP72 was elevated 24 hours after the HEAT and COM and attenuated post HST. Exercising HR, Tcore and PSI was reduced during the HST 24 hours after a heat stressor had been applied, but unaffected by a prior hypoxic exposure. Therefore the hypothesis was accepted. It was concluded that at the temperature and level of hypoxia studied, a prior exposure to exercise heat stress was beneficial when conducting subsequent acute hypoxic exercise. Study Two investigated the effect of short-term heat acclimation (STHA) on subsequent hypoxic tolerance in 16 male participants divided equally into 2 matched groups. This study also examined the response of extracellular HSP72 (eHSP72) to acute hypoxic exercise. It was hypothesized that STHA would increase basal HSP72 and that the post HST increase in HSP72 would be attenuated in this group, indicating conferred cellular tolerance. Eight males completed a HST one week before undertaking 3 consecutive days of STHA (60 min/day, 40°C, 50%peak) followed by a final HST 48-hours after the last acclimation day. The matched controls (CON) completed an identical protocol in normothermic, normoxic conditions. The initial HST induced a post exercise increase in HSP72 in both groups. HSP72 was increased after the first day of heat acclimation and unchanged in the control group. After acclimation day 2, basal HSP72 was increased from on day 1 basal values and the post exercise increase observed on day 1 was absent in the heat group. The increase in basal HSP72 persisted until the post acclimation HST for the STHA group and post exercise HSP72 was attenuated. eHSP72 increased immediately after the HST in both groups, however large inter-individual variation was evident. Mean exercising HR, Tcore and physiological strain was reduced during the HST in the STHA group, indicating that a short period of heat acclimation can improve both cellular and physiological tolerance to exercise in acute normobaric hypoxia. Study Three examined how a prior period of long term heat acclimation (LTHA) or time and absolute exercise intensity matched hypoxic acclimation (HA) affects both tolerance and performance to a HST and 16.1 km time trial (TT). Plasma hypoxia inducible 1 alpha (HIF-1α) was assessed before and after the acclimation periods as this transcription factor plays an important role in heat acclimation mediated cross tolerance. Twenty-one male participants completed ten 60-minute cycling bouts (50% Npeak) in thermoneutral, normoxic conditions (CON, 18°C, FIO2 0.209; n = 7), heated conditions (LTHA, 40°C, n = 7) or hypoxic conditions (HA, FIO2 0.14, n = 7). A HST immediately followed by a 16.1 km TT was completed one week before and 48 hours after the acclimation period. Both LTHA and HA induced increases in basal HSP72 by the end of the 10-day period. Increases in basal HSP72 occurred earlier in the acclimation period and to a greater magnitude with LTHA. Prior to the post acclimation HST both basal HSP72 and plasma HIF1-α were elevated in the LTHA and HA groups, with no changes observed in CON compared to the initial HST. Post HST mHSP72 and HIF1-α was attenuated in LTHA and HA. Mean exercising HR, Tcore and PSI were reduced in the LTHA group with no changes in these physiological variables observed in the HA or CON groups. During the TT, mean power output (MPO) was elevated at each kilometer in the HA group, leading to an improved performance after acclimation. The LTHA group produced greater power outputs between km 1 – 8 and 14-16 and consequently were faster overall compared to their pre acclimation TT. This indicates an altered pacing strategy following the LTHA period. The data suggests that, at the levels studied herein, LTHA induces a faster accumulation of basal mHSP72 over a 10-day period, occurring to a greater magnitude. This is the first study to examine the plasma HIF-1α response to both heat and hypoxic acclimation in humans. The data suggest that each environmental stressor induces an increase in resting levels of this transcription factor, however further study is required due to the large variation in response. It is not yet known whether the benefits conferred from heat to acute bouts of hypoxia would translate to more prolonged hypoxic exposures. Both the mechanisms of cross-acclimation and the effects of extended or prolonged hypoxic exposure following heat acclimation require further study. The immediate post exercise mHSP72 increase to exercise was consistently shown to be greater following a heat stress condition when compared to hypoxia. STHA induced greater increases in basal mHSP72 compared to the acute exposure, further attenuating post HST mHSP72 elevations and physiological strain. LTHA increased basal mHSP72 at a faster rate and magnitude than HA and 16.1km time trial performance improved to a similar magnitude following both heat and hypoxic acclimation It is speculated that heat acclimation mediated activation of HIF-1α may hold a key mechanistic role in the observed cross-acclimatory response. From a practical perspective, the use of heat-stress based acclimation/training programs may provide a cheaper and more effective means of preparing individuals for subsequent hypoxic exposure. Future studies should confirm these observations hold true in a hypobaric environment and establish how prior heat acclimation may impact on longer term exposures and adaptations to hypoxic environments.

Acute lower body exercise elicits adverse effects on balance performance and subsequent fall risk. However, little information exists for upper body exercise and postural sway. The series of experimental studies presented in this thesis investigated the effects of acute upper and lower body exercise on postural sway in healthy young and older adults and determined whether the acute negative effects of exercise can be removed by an improvement in training status. Chapter 4 examined the effects of maximal and submaximal (absolute and relative exercise intensities) arm crank ergometry (ACE) and cycle ergometry (CE) on postural sway in young healthy adults. Cycling elicited an immediate increase in post exercise postural sway whereas ACE did not. Chapter 5 compared the effects ACE, CE and treadmill walking (TM) on postural sway in healthy older adults. Based on the findings of Study 1, submaximal exercise was performed at the same relative intensity (50 % HRE). In agreement with Chapter 4, CE and TM elicited post exercise balance impairments lasting for ~ 10 min post exercise. ACE performed at the same relative intensity as the lower body did not elicit post exercise balance impairments in older adults. Collectively, these acute studies suggest that lower limb exercise may acutely increase fall and injury risk in the immediate period after exercise cessation. This is important because practitioners and clinicians should acknowledge that the prescription of conventional training modes might potentially elicit transient impairments in neuromuscular function. However, in this context it appears that seated exercise with the arms may not induce a significant enough stimulus to cause sensorimotor disturbance to postural stability and thus may be a safer alternative exercise mode for fall risk populations or individuals who are very sedentary. Chapter 6 examined differences in balance performance, as measured by quantitative posturography and functional balance tests, among different age groups. Measures of postural sway were able to distinguish between younger (< 60 years) and older ( > 60 years) adults whereas functional performance tests suggested that balance impairments were observed earlier ( > 50 years). This study enabled a range of tests to be determined for use with subsequent training interventions. Chapter 7 examined the effects of 6-weeks upper or lower body exercise training on postural sway. Upper and lower body training elicited similar improvements in specific (~ 25 %) and cross transfer (~ 12 %) exercise tolerance. Both modes of training elicited favourable balance adaptations. Specifically, upper body training improved mediolateral aspects of postural sway, while lower body training improved anteroposterior aspects of sway. It was proposed that an improvement in cross transfer exercise capacity after upper body exercise reduced the physiological strain experienced during CE, thus reducing post exercise balance impairments and that an increase in abdominal and trunk strength from upper body exercise training reduced sway following ACE. Importantly, ACE also elicited an improvement in functional reach distance and timed up and go test speed. Conversely, CE improved lower limb strength which elicited an improvement in lower body dynamic balance. It is likely that engaging in both upper and lower body exercise will be better that either mode of exercise alone for both health and fitness and balanced incorporated in everyday life. ACE elicits a number of significant benefits to cardiovascular fitness and balance which is important for a number of older subgroups who might have difficulty engaging in lower limb exercise, such as those who are very sedentary, those with lower limb injury/disease or undergoing surgery rehabilitation and individuals who are overweight. Specifically, ACE may offer a pathway from sedentary living to physical activity. Upper body endurance exercise can contribute to a multimodal training stimulus by eliciting favourable adaptations in fitness, functional performance and balance. Such responses are important because this type of exercise may serve a feasible and time-efficient training regime for older adults, which will likely result in improved attrition and adherence to physical activity.

Aim: To investigate the determinants of physical activity participation among primary schoolchildren in the Portsmouth area. Introduction: There is a strong body of empirical evidence to suggest a causal relationship between physical activity and health. Although many of the health-related diseases manifest themselves in adulthood, precursors or risk factors for many of these healthrelated diseases appear to be established in early childhood. Studies among children and adolescents suggest that they do not engage in physical activity of the type, duration, intensity and frequency to gain health benefits. Therefore, an understanding of the factors that influence children's physical activity would help us to comprehend how children might be encouraged to increase their participation. Very few studies have described the factors that influence physical activity participation among British primary schoolchildren; this provided the avenue of investigation for the present study. Subjects, Materials & Methods: Participants were 10-11 year-old primary schoolchildren (n=270) and their parents from three schools in the Portsmouth area. A multi-method approach was taken, combining subjective and objective methodologies, to meet the aims and objectives of the study. Three questionnaires, a self-administered report form and heart rate (HR) monitoring were chosen to obtain data about the dependent variable, physical activity participation, and twenty-five independent variables, including home media usage, school environment, socio-economic class and other parental factors. Prior to the main study a pilot study was conducted to evaluate and amend the three questionnaires. Data collection took place during one academic year. Stepwise multiple regression analysis was used to develop a 'best-fit' model identifying potential determinants of physical activity partlclpatlon among British primary schoolchildren. Results: In total 92.1% of children completed the questionnaires and self-administered report, 72.2% of children taking part in the HR monitoring provided usable data, and 54.1% of children's parents completed the parental questionnaire. The children's physical activity checklist revealed that, on average, children participated in 5.2 hours of physical activity per week (encompassing light, moderate and vigorous physical activities), or 44 minutes per day during the study period, significantly less than the stated physical activity recommendations (Health Education Authority, 1998). HR monitoring of children's physical activity participation during school break-times showed that, on average, children accumulated 15 minutes of moderate-ta-vigorous physical activity (MVPA) per day, during their school break-times. Multiple stepwise regression models were able to account for 45.3-50.0% of variance in children's physical activity behaviour. 'Community sports partiCipation', 'extracurricular sports participation' and 'having a favourite sports personality' were consistent positive correlates of physical activity among all children. Multiple regression analysis also demonstrated that the determinants of phyalcal activity participation were gender-specific. Conclusion & Recommendations: The present study provides important guidance for physical activity interventions and suggests that efforts should be concentrated on working with schools and local communities to develop opportunities for children to engage in a wide range of sport and physical activities with particular emphasis on those which are likely to encourage lifelong physical activity habits. Future studies, should build on the knowledge presented here by expanding the number and type of physical activity correlates investigated.

Children and young people’s physical activity levels are of continuing concern as the health implications for both childhood, and later on in adulthood, may be determined by their experiences of physical activity whilst growing up. Although many attempts have been made to increase physical activity levels relatively little attention has been paid to listening to the voices of children and how they experience physical activity. The publications presented in this thesis provide evidence to support the tenets of the Self Determination Theory. The results of three studies, one mixed methods, one literature review and one qualitative, have been synthesised to provide a commentary on why children engage with physical activity. The original contribution to knowledge arises from collecting evidence directly from active children and young people, many of whom have been previously disengaged from activity, to inform future interventions. For some children and young people being forced to be active is counterproductive as they associate activity as something that has to be ‘got out of the way’ or avoided where possible. If they are to take part in physical activity of their own volition it has to be enjoyable. The results of this thesis suggest that for it to be enjoyable participation should be in activities of their choice, where they feel competent and in control of their bodies whilst exercising. They want to be supported by parents and peers and respected, but not controlled, by teachers/instructors. Many children, in particular obese children, lack confidence in both their physical and social skills to interact with their peers in a physical activity setting. In providing separate activity sessions, which promote a caring and supportive climate, and a variety of non-traditional activities, children can develop competencies which enable them to enjoy being physically active. These findings need to be understood and further developed if we are to engage all young people in physical activity.

Walking groups are increasingly being set up to increase physical activity in sedentary population groups, but little is known about whether they are effective at doing so and how they work. The present thesis aims to build an evidence base of whether walking groups are effective at promoting public health and what factors account for their effectiveness. Methods: Four studies were conducted to address the overall aim. Study One: a systematic literature review with meta-analysis investigated whether interventions to promote walking in groups are effective at promoting physical activity. Study 2: a multi-perspective thematic analysis of interviews with walkers, walk leaders and walk co-ordinators, includingfollow up interviews with walkers, explored whether the needs and expectations of people who participated in walking groups were satisfied. The sample was gained from walking schemes run by Coventry City Council. Study Three: awalk-along interview study with walk leaders explored what and how environmental factors are seen to affect walking behaviours in groups. Study Four: a prospective cohort survey explored what theoretical constructspredict maintenance of attendance at walking groups in the Midlands. Results: Study One:interventions to promote walking in groups were found to be effective at promoting physical activity within efficacy studies targeting adults (d=0.42). Study Two: walkers reported that they joined walking groups to gain social and health benefits. Three months later the same walkers reported that they continued attending walking groups when their initial needs were satisfied by the other people in the group. Walk leaders and walk coordinators often acknowledged the same reasons but expressed lack of confidence to effectively address them. Study Three: walk leaders describedenvironmental factors that were important facilitators for behaviours within walking places. Lap walking places were reported to facilitate physical activity, park walking places were reported to facilitate social interactions and city centre walking places were reported to facilitate time efficient behaviours. Study Four: recovery self-efficacy and satisfaction with outcome expectancies and overall experiences within the groups were found to predict maintenance of attendance at walking groups. ix Conclusions:The results of this thesis suggest that walking groups increase physical activity. Furthermore, successful walking groups should include theory based techniques to promote behaviour change and social integration within participants. The outcomes of this thesiscan be used as an evidence base for developing, implementing and evaluating effective walking groups within the community.

Research has highlighted that adolescent girls are insufficiently active which has serious implications for their current and future health. Walking is recognised as an effective way of implementing regular, health enhancing physical activity (PA) into the daily routine of the general population and in an adolescent population walking is a convenient alternative to active play and sports participation. However it is currently difficult to promote walking and walking initiatives with adolescent girls due to lack of well-established evidence regarding both the quantity and quality of walking that might be advocated to promote PA and health. Therefore the aim of this thesis was to provide step based guidelines with regard to both the quantity and quality of walking required for health and thus inform walking interventions in adolescent girls. In order to achieve this aim four studies were undertaken. Study one explored the most appropriate way to assess walking activity, specifically whether walking on a treadmill accurately replicates walking overground in adolescent girls. Treadmill walking was found to overestimate the metabolic cost of walking in this population. This indicated that studies with the aim of exploring or promoting moderate intensity walking should focus on overground walking. Study 2 explored the quality of walking (steps·min-¹) required to achieve moderate intensity physical activity (MPA) overground. The influence of different anthropometric measures on step rate (steps·min-¹) equating to MVPA were also compared. Results suggest that a generic step rate of 120 steps·min-¹ and 7200 steps in 60 minutes may be advocated to achieve MPA in adolescent girls. However inter-individual variation in step rate associated intensity was observed and it was suggested that a step rate range based on the girl’s body mass may be beneficial for use with adolescent girls. Study 3 considered the most appropriate step measurement instrument to assess free-living walking. Five commercially available instruments (activPAL™ and pedometers; Omron HJ-720-ITC, Omron HJ-304-E, New Lifestyles NL-1000, Yamax CW-701) were compared to direct observation hand-tally step counts, during continuous (study 3a) and incidental (study 3b) walking overground. The New lifestyles NL-1000 was most consistently accurate in quantifying steps and ‘activity time’ during continuous walking, but not during incidental walking. However due to the ease of use and additional youth friendly design features, the New Lifestyles NL- 1000 was utilised in study 4. Study 4 explored the quantity of walking (steps·day-¹) required for health in adolescent girls. The results indicated that in terms of walking activity, ‘healthy’ adolescent girls do not walk significantly more in term of steps∙day⁻¹ or time spent in activity than girls classified as at ‘health risk’. Therefore specific thresholds for quantity and quality of walking required for health could not be defined for this population. Overall findings of this thesis highlight, that walking should be assessed overground with an appropriate measurement instrument. A step rate of 120 steps·min-¹ and 7200 steps in 60 minutes may be advocated to achieve MPA in adolescent girls. However further research is required to explore the relationship between walking and health in this population before we can promote an appropriate threshold of walking that is conducive to good health in adolescent girls.

This study seeks to gain a greater understanding of the process of transition and development that secondary physical education beginning teachers undergo during their pre-service year and their first year of teaching. Such an investigation is timely in the light of fundamental government reforms of current government policy that promotes a model of Initial Teacher Education (ITE) that is located in schools rather than higher education and that perceives teaching as a craft that is best learnt as an apprentice (Gove, 2010). This shift towards employment-based routes of ITE allocates universities a marginal position, which will impact on the kind of support they can provide. The research aim is to investigate beginning teachers’ journeys through complex landscapes of practice. This longitudinal study takes an interpretive approach to investigate the journey through complex landscapes of practice of thirteen physical education beginning teachers. The qualitative methodology utilises procedures associated with Charmaz’s approach to grounded theory. Data were gathered over a two year period by way of 52 semi structured, individual interviews at four points in time, namely at the start of the pre-service programme (September 2008); after the completion of their first placement school (February 2009); at the end of their pre-service programme (June 2009); and at the end of their first year of teaching (May/June 2010). The data were analysed, coded, and categorised, and explanations of theory that emerged were grounded in the data. The findings of this study have four broad sets of implications for the learning journeys of beginning teachers. Beginning teachers need time and space to be able to distance themselves from the practicalities of the school setting, which can be overwhelming in the immediacy of their demands. Both schools and universities need a shared language, so that sameness and continuity can reside in the fact that both sites are concerned with pedagogy and with the learning process of the beginning teachers High quality, structured mentoring I support is paramount in order to ameliorate the inconsistencies that occur during the learning journey. The University’s role is key in the preparation of beginning teachers to help them reconcile their personal and professional identities of who they believe they are and who they are aspiring to become. This study highlights the need for policy makers, teacher educators and schools to develop a shared understanding of the complexity of factors that influence the professional learning of beginning teachers and highlights the importance of providing beginning teachers with a (neutral) third space where they can develop the critical capacity to negotiate the competing imperatives confronting them on their learning journey. Thus this study makes a timely and important contribution to the ongoing debate around beginning teachers’ professional development and particularly in the current policy context regarding preparation for teaching.

Background: There is accumulating evidence that greater availability of neighbourhood green space is associated with better health. One mechanism proposed for this association is that green space provides a venue for, and therefore encourages, physical activity. It has also been suggested that socio-economic health inequalities may be narrower in greener areas because of the equalised opportunity for physical activity green spaces provide. However, research, exploring associations between availability of green space and physical activity has produced mixed results. Failure to account for the type and amount of physical activity which occurs specifically in green space may account for these mixed findings. This thesis therefore explored the extent to which green space is a venue for physical activity and whether this could account for better health and narrower socio-economic health inequalities in greener areas. Methods: Secondary analyses were conducted on two cross-sectional surveys of adults (16+) living in urban areas across Scotland. The first survey included individual level health, total physical activity, physical activity specifically in green space and socio-demographic characteristics. These data were matched to an objective measure of neighbourhood green space availability. The second included self-reported data on green space availability, quality, green space use, health and socio-demographic characteristics. Objective and perceived measures of green space were assessed in relation to (a) health, (b) use of green space and (c) physical activity in green space using logistic regression models. Interactions between socio-economic position and each outcome were assessed. Results: The objective availability of green space in a neighbourhood was not associated with health, total physical activity or that specifically in green space. The perceived availability and quality of green space was positively associated with more frequent use, but only perceived quality was associated with better population health. There was no evidence that socio-economic inequalities in health, use of green space or physical activity within green space were narrower in greener areas of Scotland. Conclusion: There was no evidence that physical activity specifically in green space was associated with better health or narrower socio-economic health inequalities. Further research exploring green space characteristics over and above availability, may help determine whether green space is salutogenic in Scotland.