Abstract:Many types of studies examine the in uence of selected variables on the conditional expectation of a proportion or vector of proportions, for example, market shares, rock composition, and so on. We identify four distributional categories into which such data can be put, and focus on regression models for the rst category, for proportions observed on the open interval (0, 1). For these data, we identify different speci cations used in prior research and compare these speci cations using two common samples and speci cations of the regressors. Based upon our analysis, we recommend that researchers use either a parametric regression model based upon the beta distribution or a quasi-likelihood regression model developed by Papke and Wooldridge (1997) for these data. Concerning the choice between these two regression models, we recommend that researchers use the parametric regression model unless their sample size is large enough to justify the asymptotic arguments underlying the quasi-likelihood approach.

No abstract

At the previous RESAW conference, I presented a paper as part of a panel alongside Richard Deswarte and Rowan Aust about our work on the Big UK Domain Data for the Arts and Humanities project (BUDDAH). 1 My project had investigated the presence of British disability charities on the early web and how their sites might reflect on the changing nature of disability activism following the Disability Discrimination Act 1995. 2 The subtitle of that paper was 'lessons from failure with the UK web archive'-an admission of frustration with the limitation of the author as much as (if not more than) the accessibility of the material itself. As part of my current research project at the London School of Hygiene & Tropical Medicine on the history of vaccination since the Second World War, I have once again had cause to use the UK web archive for research. This paper outlines how this research has progressed and what lessons were learned from 'lessons from failure'. My research questions were more focused and achievable, even if this may have made them less ambitious. One key conclusion, however, remains. British history from the late 1990s onwards will increasingly require knowledge of and disciplinary command over internet and web source material. I should begin by saying-or admitting-that I do not consider myself a "historian of the internet". I am currently writing a book on the history of British vaccination policy since the Second World War. 3 My Masters and PhD are in the history of medicine, and my current position focuses on the history of public health. I am a historian of postwar Britain, and as such rely heavily on documentary source materials for my work on public health policy and the ways in which this was interpreted, accepted or resisted by the general public. I also have experience with oral history interviews. My work, therefore, deals primarily with documents produced by central bodies, such as government departments, private companies, non-governmental organisations and so forth. These

No abstract