Entrepreneurship by Stephen Haslett
Responding to increasing cultural diversity and rapid changes in technology and the conduct of su... more Responding to increasing cultural diversity and rapid changes in technology and the conduct of survey research, this paper addresses the need for refined tools and improved procedures in cross-cultural and cross-national studies worldwide. Our substantive case is the level of business and social entrepreneurship in the developing world. The paper does not present results but rather proposes the outlines of a large-scale mixed mode survey experiment testing four different survey modes progressively over five Asia-Pacific developing countries to investigate mode effects within the light of business development and new technology. This study advances technology through the development of a mobile internet survey device that reduces errors from multiple sources though optimal survey design taking into account total survey error.
In 2009, a team of 38 researchers carried out the Global Entrepreneurship Monitor (GEM) survey of... more In 2009, a team of 38 researchers carried out the Global Entrepreneurship Monitor (GEM) survey of social and business entrepreneurial activity in the Kingdom of Tonga. The GEM survey estimates the proportion of the population involved in business and social start-ups and new firm creation. This paper describes not the results of the survey but rather the methodology employed. Because Tonga is spread out over a swath of the Pacific Ocean, the team faced challenges in terms of survey design, field work, logistics, and quality control. To meet the standards of rigour, we designed a 'ruggedised method' for measuring entrepreneurial activity. Countries with a teledensity threshold of less than 30% (fixed lines per 100 inhabitants) present serious challenges to survey designers. The study discusses questionnaire preparation, recruitment and training of interviewers as well as survey design issues such as sample size, response rate, sampling weights, and lessons learned. The report will assist other teams in measuring entrepreneurial activity in low-teledensity countries and provides guidelines when the study is repeated elsewhere. The study also proposes a way forward to incorporate new technologies such as tablet PCs, GPS, and GIS to address the dilemmas of measuring entrepreneurial activity in low-teledensity countries.
Papers by Stephen Haslett
In discussing the two papers in this session, “Sample size considerations for multilevel surveys”... more In discussing the two papers in this session, “Sample size considerations for multilevel surveys” by Michael P. Cohen and “Two sided coverage intervals for small proportions based on survey data ” by Philip S. Kott, Per Gosta Andersson, and Olle Nerman, I will follow a format. I will first make some general comments. Then for each of these papers (which I will consider in
Wiley StatsRef: Statistics Reference Online, 2017
Australian Journal of Zoology, 2003
Captive parma and tammar wallabies, when offered ryegrass sward in plots, visited ungrazed plots ... more Captive parma and tammar wallabies, when offered ryegrass sward in plots, visited ungrazed plots more frequently and spent longer grazing on them than on plots that were previously grazed but not visibly depleted. There was no significant difference between the two species with respect to the pattern of time spent grazing between plots, but parma wallabies tended to visit ungrazed plots more frequently than tammar wallabies. When offered carrot pieces of different dimensions both species preferred carrot pieces of greater mass, regardless of shape. Both species favoured eating in the upright (bipedal) position despite the fact that the use of this position prolonged search time between bites. Although the two species are able to forage optimally when grazing, the greater efficiency of the bipedal posture in predator avoidance may result in the selection of larger food items whenever possible.
Historical Social Research, 1990
Between 1853 and 1930 New Zealand was rapidly colonised by white settlers most of whom were Briti... more Between 1853 and 1930 New Zealand was rapidly colonised by white settlers most of whom were British immigrants. From about the late 1870s their per capita rates of minor assaults, drunkenness convictions, spirits consumption, and civil suits fell dramatically. The paper asks whether the causal structure underlying these 'disorders' was the same in every one of New Zealand's nine provincial districts. Even though New Zealand was comparatively homogenous in ethnic and cultural terms, the character of its provincial districts varied substantially in respect to other criteria such as policing, population size, level of economic development, urbanization, industrialization and so forth. To determine if the same explanatory model fits each of the nine provinces, the paper systematically applies two forms of factor analysis to a matrix of twenty two variables in each province. The paper finds that every province had a causal structure which differed in kind from that operating in the other provinces. However, these differences were not fundamental in type.
Social History, May 1, 2005
... Voter behaviour and the decline of the Liberals in Britain and New Zealand, 1911¿29: some com... more ... Voter behaviour and the decline of the Liberals in Britain and New Zealand, 1911¿29: some comparisons. Autores: Stephen Haslett, Miles Fairburn; Localización: Social history, ISSN 0307-1022, Nº. 2, 2005 , págs. 195-217. Fundación Dialnet. Acceso de usuarios registrados. ...
Springer eBooks, 1985
The calculation of degrees of freedom for hierarchical loglinear models applied to complete conti... more The calculation of degrees of freedom for hierarchical loglinear models applied to complete contingency tables is usually straightforward. The calculation for such models applied to sparse complete multiway tables however, requires checking of the marginal totals for sampling zeros in all the configurations included in the model, together with an appropriate adjustment. GENSTAT, GLIM, SAS and SPSSX do not make this adjustment. In this paper an algorithm is outlined which calculates degrees of freedom for loglinear models in such sparse tables. It has been shown that under suitable conditions, the distribution of the difference of the log likelihood ratio statistics for two nested loglinear models converges to the usual χ2 distribution, and the algorithm then provides a routine method of calculating the appropriate degrees of freedom.
Computational Statistics & Data Analysis, Mar 1, 1990
The correct method of calculating degrees of freedom for goodness-of-fit tests of lcglinear mod... more The correct method of calculating degrees of freedom for goodness-of-fit tests of lcglinear models has been a contentious topic where the multi-way table to which the model is applied contains marginal zeros due to sampling. Two methods have been proposed, one of which is ...
Linear Algebra and its Applications, Apr 1, 1996
Although updating equations for general linear models exist, they have been restricted to formula... more Although updating equations for general linear models exist, they have been restricted to formulations with uncorrelated error structure, or to models involving fixed parameters of fixed length with correlated error structure. In this paper updating equations are developed for general ...
Australian Journal of Zoology, 2003
We investigated the functional significance of molar progression and the influence of diet on the... more We investigated the functional significance of molar progression and the influence of diet on the usefulness of molar progression as an index of age in two macropodid marsupials, the tammar wallaby (Macropus eugenii), a grazing species, and the parma wallaby (Macropus parma), a browser/grazer, by exploring the relationships between the index of molar progression and several skull and tooth parameters. We also tested allometric models that related molar progression and aspects of tooth morphology to body mass. Results support the notion that molar progression in these closely related macropods results from 'mesial shift'(forward movement resulting from growth of the bones of the skull bearing the dentary, the anterior viscerocranium) rather than from 'mesial drift' (forward movement of molars relative to the anterior viscerocranium). There were no significant differences between the two species in the rate of molar progression despite differences in diet. Instead, the greater reliance of tammar wallabies on grasses was reflected in differences in their tooth morphology from that of parma wallabies. The sum of the breadths of erupted molariform teeth of tammars increased significantly faster with body mass and with length of the anterior viscerocranium than in parma wallabies and approximated a theoretical model for compensation with metabolic body mass more closely than models based on other morphological parameters. The total mesiodistal length of dentition, the mesiodistal lengths of the component teeth of the proximal molar row, and the distance between the mesial and distal lophs were all significantly lower in tammar wallabies than in parma wallabies. These differences result in tammar wallabies having greater numbers of transverse cutting edges per unit of molar tooth length, which maximises the efficiency of comminution of long grass fibres.
Special Matrices, 2019
When sample survey data with complex design (strati cation, clustering, unequal selection or incl... more When sample survey data with complex design (strati cation, clustering, unequal selection or inclusion probabilities, and weighting) are used for linear models, estimation of model parameters and their covariance matrices becomes complicated. Standard tting techniques for sample surveys either model conditional on survey design variables, or use only design weights based on inclusion probabilities essentially assuming zero error covariance between all pairs of population elements. Design properties that link two units are not used. However, if population error structure is correlated, an unbiased estimate of the linear model error covariance matrix for the sample is needed for e cient parameter estimation. By making simultaneous use of sampling structure and design-unbiased estimates of the population error covariance matrix, the paper develops best linear unbiased estimation (BLUE) type extensions to standard design-based and joint design and model based estimation methods for linear models. The analysis covers both with and without replacement sample designs. It recognises that estimation for with replacement designs requires generalized inverses when any unit is selected more than once. This and the use of Hadamard products to link sampling and population error covariance matrix properties are central topics of the paper. Model-based linear model parameter estimation is also discussed.
PubMed, Mar 1, 1998
Objective: To determine the effect of mild dehydration (5%) on expired breath H2 concentrations i... more Objective: To determine the effect of mild dehydration (5%) on expired breath H2 concentrations in dogs. Animals: 10 healthy, colony-source dogs. Procedure: Expired breath samples were collected at baseline, and every 90 minutes for 18 hours after ingestion of a test meal (commercial dog food and kibbled wheat) in fully hydrated dogs and in the same dogs when they had lost 5% of their body weight after food and water deprivation. The areas under the breath H2 concentration versus time curves (AUC) for the dehydrated and nondehydrated states were compared, using a two-stage, balanced, crossover, repeated measures technique. The number of breath samples considered to be contaminated by flatus were compared by use of a one-sided sign test. Results: Expired breath H2 concentration of dogs during the dehydration test period was significantly (P < 0.02) greater than that during the nondehydration test period. In addition, flatulence was significantly (P < 0.033) more frequent in dogs during the dehydration period. Conclusions: Dehydrated dogs have greater expired breath H2 concentration and produce more flatus after ingestion of a carbohydrate-containing meal. Considered together, these findings suggest that dehydration results in an increase in the net amount of H2 produced in the gastrointestinal tract. Clinical relevance: Care should be taken to assess the hydration status, and to correct hydration deficits of dogs prior to breath H2 testing.
Combining data sources is often seen as a panacea, having the potential to produce more to produc... more Combining data sources is often seen as a panacea, having the potential to produce more to produce cost-effective, accurate, fine-level statistics for a lower cost. This paper clarifies conditions under which Official Statistics data sources, particularly surveys and censuses or surveys and administrative sources, should and should not be combined using statistical models based on mass imputation, spatial microsimulation, and small area and domain estimation. The theoretical links between these three techniques are explored. The wider research from which this paper is a report considers the relevant literature in depth, further develops existing statistical methods, considers their application in principle to set of case studies in sociology, economics, and business, and provides guidelines for use of the three techniques based on this research.
Biological Conservation, 2021
Dispersal dynamics can determine whether animal populations recover or become extinct following d... more Dispersal dynamics can determine whether animal populations recover or become extinct following decline or disturbance, especially for species with slow life-histories that cannot replenish quickly. Palm cockatoos (Probosciger aterrimus) have one of the slowest known reproductive rates of any parrot, and they face steep decline in at least one of three populations comprising the meta-population for the species in Australia. Consequently, we estimated demographic rates and population connectivity using data from published field studies, population genetics, and vocal dialects. We then used these parameters in a population viability analysis (PVA) to predict the trajectories of the three regional populations, together with the trajectory of the meta-population. We incorporated dispersal between populations using genetic and vocal data modified by landscape permeability, whereby dispersal is limited by a major topographical barrier and non-uniform habitat. Our PVA models suggest that, while dispersal between palm cockatoo populations can reduce local population decline, this is not enough to buffer steep decline in one population with very low breeding success. The small population size and likely decrease in the meta-population of greater than 50% over three generations (49 years) supports a change of conservation status for Australian palm cockatoos from 'Vulnerable' to 'Endangered' under IUCN criteria. Our research provides an important demonstration of how PVA can be used to assess the influence of complex metapopulation scale processes on the population trajectory of species that are challenging to monitor.
IELTS research grant awarded Round 9, 2003 This study surveys all tertiary institutions in Aotear... more IELTS research grant awarded Round 9, 2003 This study surveys all tertiary institutions in Aotearoa New Zealand to understand the attitudes of decision-makers towards the English language tests used as entry to their programs. A survey of all public and private tertiary institutions in Aotearoa New Zealand was conducted to investigate the attitudes of tertiary decision-makers towards English language tests being used as entry to their programs. Results showed that the extensive and rapid changes in the tertiary environment in Aotearoa New Zealand means that it is now characterised by diversity, resulting in a wide variety of types of English language requirements for different courses. Many tertiary institutions are actively recruiting international students, and are moving away from rigid gate-keeping English language policies to more flexible 'pathways' to proficiency. There is an increased interest in the use of a range of English language assessment methods to give a fuller picture of students' abilities. IELTS is the most frequently used test, and the IELTS 'brand' is well-known, but its use in high-stakes situations means that it has taken on a symbolic value beyond its function as an indicator of language proficiency. Decision-making on English language thresholds is made in a wide variety of ways and at a wide variety of institutional levels. A number of decision-makers at tertiary institutions have said they would appreciate more information about test results from test providers. There is potential for greater liaison on language proficiency issues between course providers and external industry standard-setting bodies. Note: Aotearoa New Zealand is the bilingual name for New Zealand. It is becoming more frequently used, particularly in contexts which emphasise the inclusiveness of all New Zealanders. The M!ori name 'Aotearoa' is translated as 'Land of the long white cloud'. AUTHOR BIODATA: HILARY SMITH Hilary Smith is a language and linguistics teacher and researcher based in Palmerston North, Aotearoa New Zealand. She has lived and worked in Tonga, Papua New Guinea and the Lao PDR. Her major areas of interest include English for speakers of other languages (ESOL), language attitudes, language and education, language and settlement, and education and international development.
Journal of Interdisciplinary History, Apr 1, 2005
Although the working classes were always regarded as the "natural constituency" of the socialist ... more Although the working classes were always regarded as the "natural constituency" of the socialist and labor parties, the exact extent to which the working classes were responsible for the left's success in national elections before the advent of opinion surveys remains in question. Little is known about long-term trends in working-class voting for the left; about its spatial and other variations; about the extent to which changes in working-class party loyalties were responsible for the left's electoral breakthroughs and defeats; and about whether the mainstream left's electoral problems (especially before the 1930s) stemmed from its inability to mobilize speciªcally working-class voting support. How stable was the social composition of the left's votes, and what was the prolonged effect of such short-term contingencies as political crises, policy shifts, leadership turnovers, party tactics, and economic shocks on it? One cross-national comparison of seven European countries from about 1900 to the 1960s that attempted to answer these questions was undermined by its arbitrary deªnition of working class. The richest among the other studies are mainly conªned to two countries (Germany in the late Imperial and Weimar periods and Britain in the early twentieth century), and generally lack strong longitudinal dimensions. 1
Calcutta Statistical Association Bulletin, Mar 1, 2013
The linear mixed model, with its combination of fixed and random parameters, plays a central role... more The linear mixed model, with its combination of fixed and random parameters, plays a central role in many statistical applications. Here we review results on conditions for best linear unbiased estimates (BLUEs) of estimable functions of fixed parameters under one linear mixed model to remain BLUEs under a second model, which differs from the first in covariance structure. Without making full rank assumptions for design matrices or covariance matrices, we also review results for the conditions under which best linear unbiased predictors (BLUPs) of random parameters under the first model remain BLUPs under the second model, and for the conditions under which both BLUEs and BLUPs under the first model remain the BLUEs and BLUPs under the second. We also provide a rather generous list of related references.
Uploads
Entrepreneurship by Stephen Haslett
Papers by Stephen Haslett