Hill farming systems in Scotland are the result of long evolution and adaptation to financial, social and political changes. Farming in the hills is a major contributor to rural industry and plays an important role in the economy, environment and social cohesion of these areas. However, it is fragile and has been dependent for many decades on high and continued levels of support payments. Agricultural land managers in these hill areas are also under increasing pressure from the other land use groups whose interests lie outside farming. With recent agricultural reforms, shifts in policy orientations regarding land use and changes in support, the future role of hill farming remains uncertain. This thesis sets out to examine the role of hill farming in this context of change, by investigating how hill farmers respond to changing policy, by understanding what other interested stakeholders expect from the hills, and exploring how hill farmers may have to adapt their farming system in response to these changes and expectations. Using an adaptive conjoint analysis method, stakeholders’ expectations have been assessed. Multivariate analysis and participative research with hill farmers have also been carried out, to typify their management responses to policy changes, using the 2003 CAP reform as an example. Stakeholders’ expectations and farmers’ types were then used in a linear programming optimisation model, to explore how hill farmers can maximise their financial margins under different policy and market change scenarios, and how their motivation is a drive towards adaptation. The results suggest that whilst livestock production is identified by stakeholders as one of the most important features for the hill areas, the continuity of livestock farming in the hills is threatened, as hill farmers are strongly affected by policy and market changes. This research also shows that there is a strong diversity in hill farming systems and in hill farmers’ management styles and motivations. That, perhaps, is one of the most important factors to acknowledge when formulating policies. This research also highlights the vulnerability of hill farming businesses (especially hill cattle production) to market price volatility, policy, subsidies and support changes, making it difficult for any hill farmer to withstand these fluctuations. Although different land uses, such as planting forestry, potentially bring substantial economic benefits, they are dependent upon many other restricting factors, including government grants, and require long-term commitment before benefits are seen. Integration of different land uses and productions could be one economic option for these areas, but a set of measures, perhaps including livestock-linked subsidies that recognise the diversity of hill farmers is needed, as well as options for hill farmers to be able to provide other public goods. Unless there is stability in market prices, a policy drive towards increasing efficiency, adequate subsidies and support and more readily attractive diversification options, including the provision of other ecosystem services linked to these grassland systems, this thesis suggests that there is a danger of hill land abandonment, a further decrease in hill farming activity, leading eventually to a decline in rural areas, not unlike many other marginal hill and mountain areas in the UK and Europe.

We propose and test a new adaptive conjoint analysis method that draws on recent polyhedral “interior-point” developments in mathematical programming. The method is designed to offer accurate estimates after relatively few questions in problems involving many parameters. Each respondent’s ques-tions are adapted based upon prior answers by that respondent. The method requires computer support but can operate in both Internet and off-line environments with no noticeable delay between questions. We use Monte Carlo simulations to compare the performance of the method against a broad array of relevant benchmarks. While no method dominates in all situations, polyhedral algorithms appear to hold significant potential when (a) metric profile comparisons are more accurate than the self-explicated importance measures used in benchmark methods, (b) when respondent wear out is a concern, and (c) when product development and/or marketing teams wish to screen many features quickly. We also test hybrid methods that combine polyhedral algorithms with existing conjoint analysis methods. We close with suggestions on how polyhedral methods can be used to address other marketing problems. / Sloan School of Management and the Center for Innovation in Product Development at MIT

Nutrition labeling research suggests consumers want nutrition information on the label; however, many do not comprehend it. The purpose of Phase I was to determine the effect of: 1. two levels of nutrition label formats; 2. three levels of nutrition information load on consumers' preference for product choice using adaptive conjoint analysis. A computer interactive interview was conducted on 252 consumers in Crossroads Mall, Salt Lake City, Utah. label alternatives were printed on soup cans to realistically portray the information. The conjoint analysis compared the attributes, nutrition information format, and nutrition information load in addition to brand and price and determined how the study participants ranked choices within each of these attributes and against the other attributes. There were significant differences (p < .000) among all three mean utility values± Standard Error of the Mean (SEM) of information load, most (.300 ± .03) , more (.154 ± .02), and some (-.231 ± .03). There was no difference between graphical (.093 ± .027) and traditional (.055 ± .020) formats (p = .298). For the other attributes, there were significant differences (p < .000) among all brands, Campbell's (.590 ± .03), Private label (-.007 ± .02) Generic (-.361 ± .03) and all prices, (p < .000), low (.431 ± .03), medium (.022 ± .02), and high (-.230 ± .03). Market simulations were performed and market share was shifted from the major brand when nutrition information was added to a Private label or Generic brand. The purpose of Phase II was to determine the effect of: 1. three levels of nutrition information content load; 2. two levels of nutrition information order; 3. three levels of nutrition information format; and 4. four levels of nutrition information expression on consumers' perceptions of label usefulness in purchase decisions. The methodology was the same as Phase I. There were significant differences (p < .000) among all three mean utility values ± SEM of information load, most (.327 ± .02), more (.091 ± .02) , and some (-.213 ± .03), and between the two mean utility values ± SEM of information order, rearranged (.157 ± .03) and traditional (-.02 ± .02). Consumers significantly preferred (p < .000) the graphical format (.148 ± .02) over the graphical nutrient density (.038 ± .02) and traditional (.018 ± .03) formats. Consumers significantly preferred (p < .000) nutrition information stated in absolute numbers and percentages (.296 ± .03), versus absolute numbers only (.028 ± .03), traditional (-.026 ± .03), and percentages only (-.025 ± .03) expressions. The most useful nutrition label in a purchase decision was one that contained the most information, in a rearranged order, with a graphical format, and an absolute number and percentages expression.

Submitted by Léa Simone Friedmann dos Reys (leafried@yahoo.com.br) on 2014-06-17T04:41:33Z No. of bitstreams: 1 Tese.pdf: 25002900 bytes, checksum: 60392b62f1850ece064cd6b5a73a1acc (MD5) / Approved for entry into archive by PAMELA BELTRAN TONSA (pamela.tonsa@fgv.br) on 2014-06-17T09:28:09Z (GMT) No. of bitstreams: 1 Tese.pdf: 25002900 bytes, checksum: 60392b62f1850ece064cd6b5a73a1acc (MD5) / Made available in DSpace on 2014-06-17T12:07:51Z (GMT). No. of bitstreams: 1 Tese.pdf: 25002900 bytes, checksum: 60392b62f1850ece064cd6b5a73a1acc (MD5) Previous issue date: 2014-05-23 / Identifica-se a gestão por processos como uma abordagem eficaz para alavancar os resultados da organização. Para que isso possa ocorrer porem é necessário modificar os processos de negócio existentes. Pessoas são o único elemento comum na gestão por processos e na mudança. Para realizar os benefícios pretendidos, a ação organizacional deve focalizar seus esforços na consecução do apoio das pessoas. A revisão bibliográfica identificou mais de uma centena de fatores de influência na mudança. Esses fatores foram classificados em 4 grupos, a saber: características do indivíduo – Tipo 1, características do indivíduo – Tipo 2, características da organização e fatores de influência. Somente os identificados como fatores de influência exercem um impacto direto sobre o indivíduo. Estes fatores foram reclassificados nos 15 atributos utilizados para realizar o estudo empírico. O primeiro estudo, qualitativo, selecionou dos 15 fatores de influência no indivíduo para a mudança, os 10 mais importantes. Com esses 10 fatores, foi realizada uma pesquisa quantitativa pela internet visando identificar a importância relativa destes. Foram utilizados dois métodos para estimar a importância de cada fator para cada respondente que participou do estudo: Adaptive Conjoint Analysis (ACA) e Ordenação. Utilizando-se o método estatístico Analise de Agrupamento, foram formados, por semelhança, de 2 a 10 grupos com as importâncias relativas resultantes. Esses agrupamentos foram posteriormente confrontados com as variáveis de controle e com o fato da mudança avaliada, ter sido ou não gerada por uma demanda de tecnologia da informação (TI). Nos grupos formados pelo método ACA, há indícios de que as variáveis de controle e o fator gerador da mudança impactam o comportamento dos indivíduos. Nos grupos formados pelo método ordenação, os indícios revelam que não há impacto nem das variáveis de controle (salvo poucas exceções), nem do fator gerador da mudança. No teste não paramétrico de Friedman e nas comparações múltiplas de Dunn-Bonferroni utilizados para priorizar as importâncias calculadas no método ACA, foram encontrados dois grupos distintos. Pelo método de ordenamento, o resultado da amostra mostrou três grupos com priorização distinta dos atributos. / Process Management is identified as an effective strategy to achieve results in the organization. The achievement of desired changes requires changes to existing business processes. People are the only common element in process and change management. The achievement of expected benefits requires a focus on the efforts required to obtain the support of affected people. The literature review revealed over one hundred factors which affect change. These factors were classified in to four distinct groups, namely: Type I individual characteristics, Type II individual characteristics, organizational characteristics and influence factors. Influence factors have a direct impact on the individual. The factors identified were subsequently classified into the 15 factors or attributes used for the empirical study. The first qualitative study selected, from the 15 attributes, the 10 most important factors which influence individual change. These 10 factors were subsequently used for the internet study whose objective was to identify the relative importance of these factors. Two methods were used to estimate the importance of each factor for each individual participating in the study, namely: Adaptive Conjoint Analysis (ACA) and Ordering. The use of Cluster Analysis used similarity analysis to identify 2 to 10 groups together with resulting relative importance scores. These groupings were subsequently contrasted with control variables and whether the change, analyzed by the individual, was the result of a demand from Information Technology. For those groups formed through the use of ACA, there are indications that the control variables and the change driver have an impact on the behavior of the individual. For those groups identified though ordering analysis, the results obtained reveal, with minor exceptions, that neither the control variables nor the change driver have an impact on the behavior of the individual. The application of Friedman non parametric test and multiple comparisons of Dunn-Bonferroni for prioritizing the importance scores calculated through the ACA method, revealed the existence of two distinct groups. Ordering analysis revealed the existence of three groups with differing attribute or factor prioritization.