Silvia Biffignandi

Web Panel Representativeness

Web panels are becoming more and more popular for data collection. They present specific problems and advantages with respect to the usual modes of collection. This paper analyzes possible re-weighting adjustments for non-response in panel data. Different weighting schemes are evaluated by means of a simulation study based on real data.

Representativeness in panel surveys

ABSTRACT A panel is a set of units recruited and used in successive surveys. When the sample unit is the household, so-called R-indicators together with the comparison of distributions of certain variables to those known in the total population help to measure the representativeness of the panel. The method is applied to Understanding Society, a U.K. household longitudinal study. At each wave, under- and over-represented groups of individuals are identified. This allows the implementation of better survey designs and procedures to reduce the bias of nonresponse.

Web-Face-to-Face Mixed-Mode Design in a Longitudinal Survey: Effects on Participation Rates, Sample Composition, and Costs

Abstract Sequential mixed-mode designs are increasingly considered as an alternative to interviewer-administered data collection, allowing researchers to take advantage of the benefits of each mode. We assess the effects of the introduction of a sequential web-face-to-face mixed-mode design over three waves of a longitudinal survey in which members were previously interviewed face-to-face. Findings are reported from a large-scale randomised experiment carried out on the UK Household Longitudinal Study. No differences are found between the mixed-mode design and face-to-face design in terms of cumulative response rates and only minimal differences in terms of sample composition. On the other hand, potential cost savings are evident.

Responsive Design for Economic Data in Mixed-Mode Panels

In order to estimate socio-economic variables in a survey, time and cost constraints must be taken into account. Economic statistics survey-based data nowadays could benefit in terms of quality and cost by implementing differential design strategies thanks to the availability of registers containing auxiliary variables, and the possibility of collecting and monitoring the data collection process by using IT data storage during the process. Our study investigates the potentials of the application of responsive design strategies in the collection of data for the estimation of economic indicators. Special attention is devoted to non-response bias reduction. Some thoughts on the consequences on total survey error (TSE) are also proposed. The empirical analysis is based on data from the recruitment of a mixed-mode panel for businesses, the PAADEL panel.

Web Surveys: Methodological Problems and Research Perspectives

This paper presents a framework of current problems and related literature on Internet/web surveys. It proposes a new classification of research topics on challenging issues. Thereby it takes into account application the role that these surveys play in different research contexts (official statistics, academic research, market research). In addition critical research, open questions and trends are identified. Furthermore a specific section is devoted to bias estimation, which is a critical point in this type of survey. In particular, original bias and variance definitions are proposed.

Web Sites Communication Performance: A Methodological Approach

The Web userpsilas behavior can be affected from the Website quality i.e shortly speaking from the fact the Website allows for an efficient spread of information. Therefore there is a need of having Website quality measurements and of obtaining information on how to improve the communication performance of a Website. Usability is in principle the measurement for evaluating the ease of use of a particular website. With usability measurement two problems arise: 1) all usability evaluation tools donpsilat take into account some heuristics 2) a Website which respects usability criteria, could yet be useless since a usable tool in our opinion should allow for convergence between userpsilas expectations and ownerpsilas goals. In our paper we set up a stepwise procedure that starting from a usability evaluation, using auxiliary variables profiles Websites according to the economic profile of the owner. This procedure allows for the identification of usability criteria that take into account the ownerpsilas profile and the website characteristics. The procedure is applied in a prototypal way to the Italian municipalitiespsila Websites.

From Big Data to Information: Statistical Issues Through a Case Study

The present paper gives a short overview of the use of Big Data for statistical purposes. The introduction of different classifications of Big Data highlights the problems that arise when trying to use them in a statistical way. After that, a small-scale case study is presented by critically highlighting problems and solutions arising out of the transition from Big Data to information; it combines Census data from the Italian NSI with a telecommunication provider dataset.

Combining survey and administrative data in Italian business surveys

The main Italian enterprise surveys are shifting from a simple traditional data collection approach to a more complex one. This new approach is based on survey database management involving the recasting of many data sources (including administrative data). The reasons for utilising administrative data are: improved timeliness, increased precision and a reduction in the statistical burden. In our paper we focus on the Italian SCI survey (Sistema dei Conti delle Imprese, i.e., Businesses accounts survey). A number of innovations have been introduced into this survey over the last few years and remain ongoing. Since 2005, final economic estimates are obtained combining various data sources, primarily administrative data. The integration procedure involves a number of methodological solutions. In this paper we deal with the problem of non-response, particularly unit non-response. At first methodological issues, research and applicative trends in the NSI (National Statistical Institute) are briefly reviewed. Afterwards alternative means of estimating business data using administrative records and integrating sources are applied to the SCI survey data. The integration procedure is presented and its impact on the improvement of final data quality is verified.

Technology convergence on telecommunications systems integration

This work introduces a technological integration applied to the telecommunications equipments of the given case study at the University of Bergamo in Italy, where research is being developed on the existing technological instruments and on their costs. By creating the operational logistics for these instruments, it was compared with the latest technological products and changes were proposed with regard to the devices or in the way they are being used, in order to guarantee a significant cost reduction for the university. This work also introduces a model for the evaluation of more general cases for medium and large companies. Starting from market data an algorithm is proposed to evaluate the best product marketing and technological services to choose.