Bernard Baffour

Weighting Strategies for Combining Data from Dual-Frame Telephone Surveys: Emerging Evidence from Australia

Abstract Until quite recently, telephone surveys have typically relied on landline telephone numbers. However, with the increasing popularity and affordability of mobile phones, there has been a surge in households that do not have landline connections. Additionally, there has been a decline in the response rates and population coverage of landline telephone surveys, creating a challenge to collecting representative social data. Dual-frame telephone surveys that use both landline and mobile phone sampling frames can overcome the incompleteness of landline-only telephone sampling. However, surveying mobile phone users introduces new complexities in sampling, nonresponse measurement and statistical weighting. This article examines these issues and illustrates the consequences of failing to include mobile-phone-only users in telephone surveys using data from Australia. Results show that there are significant differences in estimates of populations’ characteristics when using information solely from the landline or mobile telephone sample. These biases in the population estimates are significantly reduced when data from the mobile and landline samples are combined and appropriate dual-frame survey estimators are used. The optimal choice of a dual-frame estimation strategy depends on the availability of good-quality information that can account for the differential patterns of nonresponse by frame.

Ethnic Differences in the Quality of the Interview Process and Implications for Survey Analysis: The Case of Indigenous Australians

Comparable survey data on Indigenous and non-Indigenous Australians are highly sought after by policymakers to inform policies aimed at closing ethnic socio-economic gaps. However, collection of such data is compromised by group differences in socio-economic status and cultural norms. We use data from the Household, Income and Labour Dynamics in Australia Survey and multiple-membership multilevel regression models that allow for individual and interviewer effects to examine differences between Indigenous and non-Indigenous Australians in approximate measures of the quality of the interview process. We find that there are both direct and indirect ethnic effects on different dimensions of interview process quality, with Indigenous Australians faring worse than non-Indigenous Australians in all outcomes ceteris paribus . This indicates that nationwide surveys must feature interview protocols that are sensitive to the needs and culture of Indigenous respondents to improve the quality of the survey information gathered from this subpopulation.

An investigation of triple system estimators in censuses

The value of a census cannot be overstated, given that no other data resource provides such detailed information about the population. Further, censuses are often the only historical data source to map out change over time due the consistency of questions asked. However, it is often the most expensive undertaking-other than going to war-that a country embarks on. Countries are thus seeking more cost-effective alternatives. This paper details some exploratory research into one such alternative, based on capture-recapture methods. Capture recapture methods have been used for population estimation for decades, but the focus has been on dual system estimation. Dual system measurement of the population has been criticized for its reliance on the independence assumption between the two systems. This assumption is untestable, and failure introduces bias into the estimates of the population. The most logical improvement of dual system estimation is triple system estimation. In this paper, a simulation study is carried out to compare the performance of different dual and triple system estimators of the population size under various dependency scenarios. Performance is explored through both the bias and variability. The study shows that the dual system estimator copes well with dependence, provided the coverage of both lists are reasonably high. In addition, although the triple system estimators yield less biased estimates of the population, the dual system estimator is shown to be robust enough to cope with low levels of dependence.

An evaluation of census quality

The primary purpose of a census is to provide accurate estimates of a countrys population. These underpin a myriad of key planning decisions in between consecutive censuses at both the local and national level. In addition, reliable census data is needed at an international level. For example, the UN requires member states to take at least a census every ten years and EU member states have a statutory requirement to provide comprehensive population and housing data at regular defined periods. As such censuses have a pivotal role to play within official statistics. There are, however, different modes in which a country can conduct a census. Furthermore, despite all efforts, it is inevitable that the census cannot be perfect. Therefore, there is the need to measure the quality of the data produced in a census. The measurement of quality within a census is not an easy undertaking, mainly because there is currently no standard method of quality assessment that applies to all census methodologies. Thus this paper aims to bridge this gap, by examining quality assessment with regards to population censuses. It defines census quality, and then discusses the assessment of quality for different types of censuses.

Profiling the mobile-only population in Australia: insights from the Australian National Health Survey

Background: The Australian population that relies on mobile phones exclusively has increased from 5% in 2005 to 29% in 2014. Failing to include this mobile‐only population leads to a potential bias in estimates from landline‐based telephone surveys. This paper considers the impacts on selected health prevalence estimates with and without the mobile‐only population.

Subsequent Migration of Immigrants Within Australia, 1981–2016

Australia is a major immigration country and immigrants currently represent around 28% of the total population. The aim of this research is to understand the long-term consequences of this immigration and, particularly, how migrants respond to opportunities within the country after arriving through the process of subsequent (internal) migration. The focus is on major immigrant groups in Australia, including persons born in the United Kingdom, New Zealand, China and India, and how their patterns differ from persons born in Australia. To conduct this analysis, we have gathered data for a 35-year period based on quinquennial census data. We also obtained birthplace-specific mortality data for constructing multiregional life tables for the immigrant populations. Subsequent migration is important for understanding population redistribution, and the relative attractiveness of destinations within host countries. Our results highlight the importance of subsequent migration and the diversity of migration behaviours amongst different immigrant groups in the context of overall declines in internal migration since 1981.

Polling bias and undecided voter allocations: US presidential elections, 2004-2016

Accounting for undecided and uncertain voters is a challenging issue for predicting election results from public opinion polls. Undecided voters typify the uncertainty of swing voters in polls but are often ignored or allocated to each candidate in a simplistic manner. Historically this has been adequate because first, the undecided tend to settle on a candidate as the election day draws closer, and second, they are comparatively small enough to assume that the undecided voters do not affect the relative proportions of the decided voters. These assumptions are used by poll authors and meta-poll analysts, but in the presence of high numbers of undecided voters these static rules may bias election predictions. In this paper, we examine the effect of undecided voters in the 2016 US presidential election. This election was unique in that a) there was a relatively high number of undecided voters and b) the major party candidates had high unfavorability ratings. We draw on psychological theories of decision making such as decision field theory and prospect theory to explain the link between candidate unfavorability and voter indecisiveness, and to describe how these factors likely contributed to a systematic bias in polling. We then show that the allocation of undecided voters in the 2016 election biased polls and meta-polls in a manner consistent with these theories. These findings imply that, given the increasing number of undecided voters in recent elections, it will be important to take into account the underlying psychology of voting when making predictions about elections.

The Sources and Diversity of Immigrant Population Change in Australia, 1981–2011

Australia has one of the largest percentages of immigrant populations in the developed world with a highly regulated system of immigration control and regular censuses to track their changes over time. However, the ability to explain the population change through the demographic components of immigration, emigration, and death by age and sex is complicated because of differences in measurement and sources of information. In this article, we explore three methods for reconciling the demographic accounts from 1981 to 2011 for the Australia-born and 18 foreign-born population groups. We then describe how the immigrant populations have changed and what has contributed most to that change. We find that the sources of immigrant population change have varied considerably by age, sex, country of birth, and period of immigration. Immigrants from Europe are currently the oldest and slowest-growing populations, whereas those from elsewhere are growing rapidly and exhibit relatively young population age structures. Studying these patterns over time helps us to understand the nature of international migration and its long-term contributions to population change and composition.