Aim: The purpose of this research was to investigate the effects of different figure skating boots on kinetic and kinematic properties of the landing impact. The focus was on the vertical ground reaction force (VGRF) and kinematic differences between the examined boots duringthe landing impact as well as the material and construction design differences between the boots. The hypothesis was that different figure skating boots (Graf Edmonton and Edea Concerto) differently affect the VGRF and kinematics during the landing impact so some figure skating boots may reduce the impact at landing and could therefore possibly prevent injuries. Method: Six subjects each simulated figure skating landings wearing two different types ofskates from two different heights, on artificial ice. Differences were investigated between the boots in angular position of individual body segments and in the VGRF that acts between the skate and a force platform underneath a plastic ice surface, as well as the force which acts between the skate and the foot. The landings were examined with a Pedar-X in-shoe pressure measuring system, a Kistler force plate and a Qualysis motion capture system. The obtained values of each variable were statistically compared between the boots using SPSS. Results: Statistically significant (p≤0.05) differences between the boots were seen in the degree of right ankle and knee joint flexion during the initial landing impact, in the degree of right ankle joint peak flexion, as well as in the force magnitude experienced by the right foot during the landing impacts as well as for the area under the max force-time curve. Furthermore, differences were seen in the construction design, materials, sizes of the outsole and in the stiffness. Conclusions: This research suggests there are differences in lower extremity VGRF and kinematic effects of landing impact when using different boots. It also suggests that some figure skating boots reduce the loading on the human body during the landing impacts more than others and therefore could help in preventing injuries.

The aim of this thesis was to investigate the effect that consuming pre- and post exercise meals, varying in GI, would have on physiological responses and subsequent endurance performance. Study 1 investigated the effects of consuming either a low or high GI meal 45 min prior to exercise on cycling TT performance. It was observed that consuming a low GI meal 45 minutes prior to performing a 40k TT, resulted in a significantly improved performance (p = 0.009) compared to the ingestion of an isocaloric high GI pre- exercise meal (93 ± 8 min vs. 96 ±7 min for low and high GI, respectively). The results suggested that the ingestion of the low GI food led to an increase in the availability of CHO and greater CHO oxidation throughout the exercise period and subsequently spared limited muscle and liver glycogen stores. Study 2 investigated the effects of low and high GI 24 h diets, following glycogen depleting exercise, on TT performance the following day. No difference was observed in TT performance following low (90.7 ±11.1 min) and high (93.5 ± 9.29 min) GI, 24 h recovery diets (p = 0.35). This study concluded that provided the amount of CHO consumed during the recovery period is sufficient enough to replenish depleted muscle glycogen stores, the GI of the recovery diet offers no further benefit to performance. Study 3 investigated the effects of high and low GI recovery meals on TT performance following a short term recovery period from a glycogen depleting exercise. No significant difference was observed in TT performance between the low (90.7 ± 11.1 min) and high GI (93.5 ± 9.3 min) trials (t = 1.1; p = 0.35). This study concluded that if the following exercise bout is of short duration, and not long enough to challenge glycogen stores, provided the amount of CHO is sufficient during the recovery period, the GI of short- term recovery diets has no influence on subsequent exercise performance. In conclusion, the findings presented in this thesis should contribute to and support previous research within this area and help to contribute to the body of knowledge through greater ecological validity.

Obesity is a heterogeneous, complex, and chronic condition with large individual differences. Lifestyle modification has been widely acknowledged as the primary treatment for obesity. Objective – This PhD examined the effects of a non-dieting exercise-based lifestyle intervention programme (e.g. no calorie-restriction) using the tenets of the self-determination theory (SDT; Deci & Ryan, 1985b) to inform intervention decisions and identify individual differences (e.g. SDT was used to identify self-regulatory profiles), on physical and metabolic fitness, and psychological well-being among premenopausal, clinically obese women. The programme titled WHEEL focused on health outcomes rather than weight loss. Design – A randomised, delayed start RCT feasibility study. This longitudinal study ran for one year in two phases: a) 12 weeks of intensive intervention and b) a 40-week maintenance phase. Setting – Free living, general community setting. Participants – 62 predominantly white Caucasian (97%), clinically obese (BMI ≥30kg/m2), pre-menopausal women with a mean age of 40.2 years, free of obesity-related diseases and fit to for exercise were randomly assigned to a non-dieting lifestyle intervention group (n = 31) or waiting list control (n = 31). Intervention – Exercise: four hours of exercise per week chosen from the following options: Tai Chi, Circuit classes x 2; and Aqua aerobics. Participants were required to complete two sessions in a WHEEL class, but were encouraged to do all four. If this was not possible they had to agree the exercise of their choice with EB who checked their plan against the FITT principle of exercise. The tenets of SDT, namely autonomy, competence building, and relatedness were used to inform the design of exercise sessions. Autonomy: participants chose their own exercise programme structure. Flexibility within exercise sessions allowed for matching activities to participants’ current state of fitness. Those with high functional limitations were given alternative, seated exercises. Relatedness was fostered in different ways: 1) Outside of WHEEL:participants were encouraged to share their weight related experiences with each other. Routes to exercise venues were planned and they were encouraged to have a car-sharing scheme; and participants organised various charity walks for the group on their own accord. 2) Within WHEEL: participants generally worked in pairs whilst exercising in a group setting. After the initial 12-week intervention phase they were also allowed to invite a female friend or family member to join them in the classes. Competence building: participants were taught exercise skills; including naming and executing each exercise routine correctly, with a view of them joining ‘regular’ classes in the future. Furthermore, they were taught to take their own pulse and monitor their heart rate throughout sessions. The psycho-educational classes targeting dieting behaviours and eating regulation using Brief Cognitive Behavioural Therapy (CBT) techniques: a two one-hour session per week for three weeks, delivered in the 12-week intervention phase, challenging maladaptive eating behaviours, whilst educating participants about food labels and food choices. Educational Sessions: one per week for 12 weeks on physiological and psychological mechanisms of exercise and dieting (e.g. dangers of weight cycling due to dieting). Social Support: follow-up phone calls if two weeks of exercise sessions were missed. Adherence: attrition and attendance were monitored. Data Analysis – Mixed Method: sequential QUAN-QUAL data analyses. QUAN: intention to treat analysis, repeated measures analysis of variance, regression, and correlations. QUAL: analytic induction analysis using the QSR*NVivo qualitative data analysis software. Outcome measures at baseline, 12 weeks, and 52 weeks. QUAN Psychological Instruments: General Causality Orientation Scale (GCOS; Deci & Ryan, 1985b), General Well-Being Schedule (GWB; Dupuy, 1977 & 1978), Perceived Stress Scale (PSS; Cohen, Kamarck, & Mermelstein, 1983), Self-Perception Profile (SPP; Messer & Harter, 1986), State Self-Esteem Scale (SSES; Heatherton & Polivy, 1991), Multidimensional Health Locus of Control Scales (Form C) (MHLC; Wallston,Wallston, & DeVellis, 1978), and Social Support for Exercise Scale (SSSE; Fox & Dirkin, 1992). QUAN Physiology measures: metabolic and cardio-respiratory fitness. QUAL: 62 weight history interviews at baseline with 36 follow-up interviews, including 12 drop-outs. The semi-structured interviews explored participants’ history and prevalence of self-reported dieting and eating behaviours, assessed weight cycling prevalence and development of weight status up to baseline, investigated previous exercise history and skills, perceived health status and difficulties with physical activity including barriers, and examined motivation, goals, and expectation for WHEEL from the personal point of view and from the programme’s. The follow-up interviews at 52 weeks explored difficulties with exercise behaviour change, and quality of life. Results – Baseline: participants showed high levels of psychopathology indexed by high levels of stress, low levels of general well-being (81.8% in severely distressed category of the General Well-Being Schedule) and self-perceptions (e.g. self-esteem, body image), low autonomy and high impersonal orientation, and problems with emotional eating (78%) and dieting (86%). Also, participants had poor fitness levels (< 10% percentile for women) and metabolic profile with 50% of the participants meeting the metabolic syndrome criteria. Participants had unrealistic expectations (35% expected weight loss) and low exercise self-efficacy, low confidence in their ability to achieve, and a number of problems associated with their excess body weight. Finally, participants experienced societal prejudice in various aspects of their lives (e.g. healthcare, work). RCT phase: significant improvements in psychological functioning indexed by significant improvement in well-being (29.9% improvement in total score of GWB Schedule and all its subscales), self-perceptions (athletic, appearance, global self-worth scales of the SPP), and perceived received social support (reducing significantly the discrepancy between need for support and received support). In addition, cardiorespiratory fitness improved significantly in the intervention group (9.3% increase adjusted VO2, ml-kg- 1 min-1; 7.8% absoluteVO2, ml-min-1) as compared to controls (4% reduction adjusted VO2, ml-kg-1 min-1 & 3.2% absoluteVO2, ml-min-1). All these changes took place despite the absence of significant weight loss. Maintenance: those who continued the programme showed improved psychological functioning at 12. The participants showed significant improvements in general well-being: the average value at this stage was 74.4 (±16.6) bringing the group as a whole into the positive well-being category. Most subscales of the SPP showed significant improvements from baseline to 12 months and the discrepancy between needed and perceived provided social support for listening, information, and challenge support for exercise narrowed significantly. In support of SDT, participants felt more autonomous and more in control of their own destiny. Conclusion - Although there was a significant dropout in the study (60%) the present intervention was successful in bringing about behavioural change in those who stayed in the programme. Both the QUAN and QUAL results provided strong support for the improved psychological profile of participants in the absence of significant weight changes. Reasons for dropout included: research design, facilities, and personal. Although the study was not without limitations the underlying philosophy adopted was rarely questioned and would provide a basis for definitive RCT trail.

The overall purpose of this research programme was to examine how international adolescent golfers cope with performance related stressors. As such, three interrelated studies were designed to pursue this purpose. The purpose of Study 1 was to examine instances when international adolescent golfers' coped effectively and ineffectively with performance-related stressors during competition. Strategies associated with effective coping were rationalizing, re-appraising, blocking, positive self-talk, following a routine, breathing exercises, physical relaxation, and seeking on-course social support. Alternatively, different types of coping responses (trying too hard, speeding up, routine changes, negative thoughts, lack of coping) were associated with ineffective coping. The purpose of Study 2 was to examine stressors, coping strategies, and perceived coping effectiveness among elite adolescent golfers longitudinally over 31 days. Overall, most frequently-cited stressors were making a physical error and making a mental error. Coping strategies that served a problem-focused coping function were cited more often than those which served an emotion-focused or avoidance function. Although mean coping effectiveness remained stable over time, considerable fluctuations in the effectiveness of coping strategies used to manage specific stressors were observed. The purpose of study 3 was to identify and examine adolescent golfers' stress appraisals and coping attempts during golf performance. Stress appraisals appeared to be related to the participants' performance goals, and an array of different coping attempts was deployed to manage apparently similar stressor-appraisals The findings presented in this research programme suggest that adolescent golfers use a plethora of different coping strategies during golf to cope with performance related stressors. The types of coping strategies utilised by the participants were very similar throughout all three of the studies ranging from blocking to positive appraisal.

‘While performance analysts have been increasingly employed in many elite level sporting organisations, this development has arguably not been matched with a comparable level of critical scholarship addressing how performance analysts experience, understand and practice within a variety of organisational cultures’ (Huggan, Nelson & Potrac, 2015, p. 505). Specifically, the micro-political, emotional and identity experiences faced by neopphyte performance analysts as they transition into an elite rugby league coaching environment have not yet been considered. To partially address this situation, the current thesis provides a novel insight into my emotional, identity and micro-political experiences, as I endeavoured to transition into the coaching team for the first time. Data for this investigation was gathered in the form of a daily journal, based on my experiences at the Club, the University and in my home life. These journals were also discussed during monthy supervisory meetings, in order to further understand my experiences. Several interrelated themes emerged from the resulting narrative and were principally understood in relation to the work of Kelchtermans (e.g. Kelchtermans, 2009; Kelchtermans & Ballett, 2002a, 2002b), Goffman (1959, 1963, 1969), Hochschild (1979, 1983), Burke and Stets (2009) and Stryker (1980). I contend that the inherent structural vulnerability of my neophyte performance analyst position, as well as my determination to protect and advance my career meant that I had to learn to act micro-politically. I also assert that the multiple identities I possessed during this time acted cooperatively and conflictingly at different times, ultimately resulting in me not wanting to remain within my professional identity. It is hoped that through presenting the ambiguity, pathos and dynamic nature of practicing performance analysis in elite rugby league, a more grounded understanding of this topic area can be obtained.

The primary aim of the current project was to investigate the isokinetic muscular strength and performance of elite male youth footballers, and the relationships with age, seasonal variation and injury. A secondary aim was to use the information gathered to target muscle strain injury prevention strategies to particular age groups and times, and evaluate the effect. The primary aim was achieved by establishing normative patterns for muscular strength and performance of elite male youth footballers (grouped according to chronological and biological age) across a competitive season of youth football in Chapters Four and Five. Isokinetic muscular strength (characterised by peak torque (PT) and peak torque relative to body weight (PTBW)) of the hamstrings (H) and quadriceps (Q) using both concentric (CQ, CH) and eccentric muscle actions (EH) was evaluated. Muscular performance of the same muscle groups (characterised by H:Q ratios (conventional (CHQ) functional (FHQ)), asymmetry (dominant (dom):non dominant (ndom) leg ratios (e.g. CQ:CQ)), and angle of peak torque (AoPT)) was also investigated which necessitated an isokinetic speed of 60 °/s. Isokinetic evaluation was completed three times over the course of a regular playing season (start of season (SS) mid season (MS) and end of season (ES)). Participants were grouped according to chronological age (n=152, under 12 (U12) - under 18 (U18)) and biological maturation (according to Pubertal Development Scale (PDS 1 - 5) n=134). Forty seven participants completed SS, MS and ES isokinetic evaluation. Bilateral isokinetic evaluation consisted of five maximal repetitions of CQ and CH, followed by five repetitions of EH, leg dominance was counter-balanced. Repetitions two-four were used to calculate PT, PTBW, dom:ndom and AoPT for CQ, CH and EH, CHQ and FHQ; these measures were compared across chronological and biological age groups using a mixed model ANOVA. Dom:ndom CQ, CH and EH were compared across chronological and biological age groups using a one way ANOVA, while the relationship between AoPT and PT/PTBW was considered using a Pearson’s correlation. Additionally, the relationship between chronological and biological age, and PT/PTBW was investigated using a mixed model ANOVA within PDS group three. For analysis of seasonal variation a mixed model ANOVA was applied for all isokinetic measurements which considered time (SS, MS, ES), leg dominance (dom, ndom) and age group (U12 -U15) with a further mixed model ANOVA performed on CQ:CQ, CH:CH and EH:EH. Where appropriate SIDAK corrections were applied and the level of significance was accepted at p≤0.05. The main findings were that youth footballers did not increase their PT and PTBW EH in-line with CQ and CH as chronological and biological ageing progressed, this lead to a significant FHQ imbalance at U18. Dom:ndom CH comparisons identified that the chronologically younger and biologically less developed groups displayed a significantly stronger dom leg which may be explained through the concepts of skill acquisition and trainability. Biological age was not found to exert any additional effect over and above that of chronological ageing as significant differences in muscle strength still existed according to chronological age group within PDS group three. Additionally, AoPT EH and PT EH were found to be significantly negatively correlated on both legs which supported a potential mechanism for non contact hamstring muscle strain injury during running. Analysis of seasonal variation revealed that all PTBW measures showed a MS decrease. This may be related to breaks in normal training activity and links appropriately to times of peak injury incidence highlighted in youth football. In order to achieve the secondary aim of the current project Chapters Four, Five and Six investigated the relationship between isokinetic muscular strength and performance, muscle strain injury of the thigh, and injury risk attenuation. A retrospective and prospective injury audit was undertaken for the elite male youth football participants. For the retrospective approach participants were grouped according to chronological age (n=147) or biological age (n=128) and indicated using a self-report injury form their history (ever, (Hx)) or recent history (12 months, (Hx12)) of hamstring, quadriceps and adductor injuries. Approximately each player had an Hx of muscle strain injury and 0.56-0.59 of players had an Hx12. The hamstrings were the most commonly injured muscle group and the prevalence of muscle strain injury Hx and Hx12 increased with chronological and biological age. The prospective audit (n=50) identified that 0.16 of players sustained a muscle strain injury during the season, 0.08 of these being to the hamstrings. Between group comparisons (one way ANOVA with SIDAK correction) were also performed to investigate the difference in isokinetic measures between those participants who had an Hx12 of muscle strain injury and those who did not. It was discovered that for Hx12 of an injury to the dom hamstrings the injured group had less PTBW CH and EH on the dom leg. The injured group also had more inner range AoPT CH. These findings linked appropriately to the reported mechanisms and risk factors for hamstring injury but the exact direction of cause and effect could not be established. To this end a logistic regression analysis was undertaken in an attempt to predict which group (injured vs. non injured the 50 participants would belong to, using evidenced based risk factors in the experimental model. No predictive relationship between risk factors (including altered isokinetic muscular strength and performance) could be established. The information regarding the relationship between injury and muscular strength and performance may highlight a role for isokinetic screening to ensure adequate rehabilitation from injury. Injury risk attenuation strategies were investigated through an exercise intervention using the U18 age group following a break from football activity. The participants were split based on their FHQ at initial isokinetic evaluation (via odd and even placing) to form control (n=8) and intervention groups (n=8). Isokinetic evaluation was conducted as previously outlined and the exercise intervention targeted the hamstrings. Only six of the control group and seven of the intervention group completed the study and were compared using a mixed model ANOVA. Results showed that the intervention group were not significantly different to the control group post intervention for any of the isokinetic muscular strength and performance measures, though both groups significantly improved over time for the ndom leg CHQ and PTBW EH, and FHQ improved for both legs. Contamination of the control group may explain the lack of significant difference between groups. However, the exercise intervention was not targeted to individuals who displayed prior alterations to isokinetic muscular strength and performance, and this approach was discussed using the results of one member of the intervention group. In summary, the current project achieved the stated aims by discovering normative patterns of isokinetic muscular strength and performance according to age and seasonal variation. Injury risk attenuation strategies were targeted appropriately to the U18 age group following a break from football activity. However, the applied evidence based exercise may have been more effective if targeted to ‘risk’ after isokinetic screening.

Physical activity is positively related to a number of health benefits that influence morbidity and mortality during childhood, adolescence and adulthood. However, an epidemic of physical inactivity is quickly expanding worldwide and particularly affecting the Greek population. Early life periods and especially transitional ones leading to young adulthood are considered critical to intervene to help people adopt and maintain an active lifestyle. Well-designed longitudinal interventions are recommended for these ages. The main objective of this study was to design, implement and assess an intervention to help students adopt a more active profile according to the Transtheoretical model. This theory was selected due to its practicality and adaptability. The intervention materials consisted of a set of five printed manuals based upon the Transtheoretical model and encouraging physical activity. The study design was quasi-experimental (n=665, mean age=15.8 years, 57% girls) with a stratified assignment of the intervention (nInt=263) and control group (nCon=402). The intervention consisted of the administration of one printed manual to each student according to his/her current stage and its use for the next four months. Greek secondary students were measured longitudinally in the course of three years extending from two years before their graduation until one year after their graduation. The first two measurements were performed in the second grade of Lyceum (Greek high school) one just before and one just after the intervention. The last two measurements were conducted one year after and two years after the intervention. The research questionnaires measured stages of change, processes of change, decisional balance and self-efficacy, which are the main components of the Transtheoretical model. These instruments assisted firstly with the implementation and secondly with the assessment of the intervention. The research hypotheses examined the various intervention effects. The main analysis of the stage data was performed with latent transition analysis, which was considered as appropriate and advantageous. The latent stage results revealed positive intervention effects in the short-term, which were neutralised in the mid- and long-term. A comparison of the observed stage data pre- and post-intervention confirmed that in the short-term the intervention had successfully helped more students to progress and fewer students to regress along the stages of change continuum compared to the control group. Regarding self-efficacy, decisional balance and processes of change, within-group longitudinal comparisons of the observed data disclosed positive comparative short-term effects. In general, these effects were also reversed or neutralised in the midterm and remained neutral in the long-term. In most cases the above-mentioned trends of the whole sample were also confirmed for each gender separately making the intervention successful only in the short-term. Several shortcomings identified in the literature were addressed by the current study by implementing a longitudinal design, conducting a long-term investigation of the intervention effects and specifically adapting and validating the research instruments for the studied population. The “less is more” approach encapsulates the philosophy behind this intervention. In fact, the resources used were kept in a minimum regarding students’ time and schools’ involvement. Together with the easiness of the administration of the intervention contributed to the potential of being easily generalisable to wider populations. Additionally, the development and implementation of the Greek adolescent stages of change manuals was a pioneer work for Greece. It is recommended that a number of successive interventions be implemented to accomplish a longer duration of positive results. Another recommendation was to expand the public impact of this intervention by attempting it on a larger, even national scale and in different settings. Finally, the positive conclusions of the current study confirmed its success in helping young people adopt and maintain an active lifestyle and also it provided similar future studies with validated tools and added experience to continue in the search for more efficient PA interventions.

The congested fixture schedules in elite soccer leagues around the world has bought the issue of recovery between games and subsequent performance to the fore in soccer related research. Studies have described the time-course of recovery for numerous biochemical and physiological measures of performance, fatigue and recovery from match-play. However, the research also suggests that there is individual variation in the external load both between players and between matches. The external load measured as distance in match-play has been shown to vary by ~30% between games. However it is the internal training load that will determine the magnitude of the physiological responses on an individual basis. Therefore the major aim of this thesis was to examine the dose-response relationships between measures of training load and the physiological and biochemical responses used as markers of recovery from match-play. The thesis also assessed the relationships between these proposed markers of recovery and soccer specific performance. In meeting the aims of the thesis a number of preliminary studies were conducted. The study in section 3 assesses the extent of fixture congestion in the English Premier League. The results showed over 30% of games for the most successful teams are played with 3 days recovery time, justifying the need for investigating recovery from soccer match-play. Given the variation in soccer match-play section 4 examines the reliability and validity of the modified BEAST90 soccer simulation. A measure of performance with less variance would allow changes in soccer specific performance to be identified with greater certainty in section 7. Section 5 assesses the influence of intermittent exercise on the blood lactate response. Given that the new iTRIMP method of measuring internal training load weights exertion with the blood lactate response it was important to assess the influence exercise mode may have on the calculation of internal training load. The results showed that at higher intensities intermittent exercise produced significantly higher blood lactate responses. Section 6 assesses the dose-response relationships between training and fitness using numerous measures of internal training load over a 6 week training period. The results showed only the iTRIMP method showed a significant relationship with changes in fitness. Section 7 assesses the dose-response relationships between exertion in soccer match-play and various physiological measures of fatigue and recovery. The relationships between these measures and changes in soccer specific performance were also assessed. Finally the internal and external load were integrated and the relationships of this ratio assessed with measures of fitness and performance. The results showed that changes in any of the physiological and biochemical measures used to assess recovery did not relate to changes in performance with the exception of testosterone which showed significant positive relationships with changes in distance covered from the 1st trial of the modified BEAST protocol to 2nd. Testosterone also was the only measure to show a significant relationship during the recovery period with any measure of training load (sRPE). Finally, the novel findings of this thesis is the relationships between the integrated ratio’s of internal and external training load with measures of aerobic fitness is also presented in section 7. The studies provided in this thesis have made a major contribution in demonstrating how data that is routinely collected at elite levels of soccer can be used more appropriately. It has also shown limitations of some the methods currently employed to measure training load. Furthermore changes in many of the markers used to assess recovery of soccer players do not seem to relate to changes in soccer specific performance. This may point to a change in paradigm which is required in both research and practice.

Following surgery, amputees must re-learn how to perform various movement tasks using altered lower limb mechanics. In order to optimise the process of re-learning these tasks and inform rehabilitation practice, an understanding of the longitudinal adaptations that occur both during and following a period of rehabilitation must be established. Scientific literature has reported the biomechanical, balance and quality of life (QOL) characteristics of transtibial amputees. However, no studies to date have outlined how these characteristics develop over time. The aim of this thesis, therefore, was to investigate the longitudinal changes that occurred in unilateral transtibial amputee movement, balance and QOL from their first treatments following amputation up to six months post-discharge from rehabilitation. Studies one and two assessed the kinematic and psychological adaptations that occurred during the rehabilitation of 15 unilateral transtibial amputees. The amputees were randomly allocated into two groups, differing by early walking aid (EWA) used. One group used the Amputee Mobility Aid (AMA), which incorporated an articulation at the knee joint. The other group used the Pneumatic Post-Amputation Mobility Aid (PPAM) with no articulation at the knee joint. Amputee�s gait and quality of life (QOL) were assessed at five standardised time points using three-dimensional motion capture and the SF-36 questionnaire, respectively. Overall, amputee�s gait improved with walking velocity increasing over time (p

Thesis (Ph.D.)-University of Durban-Westville, 1992.