A standard format for representing richly encoded qualitative data is necessary because: it ensures consistency across datasets; it supports the development of common web-based publishing and search tools; and it facilitates annotated data interchange and comparison among data collections. Importantly, it should enable data and linked products to be imported and exported directly into and out of CAQDAS packages, avoiding the reliance on just a single product, and offering the opportunity to share analytic workings outside the confines of any particular software. The DExT project is developing, refining and testing models for data exchange for qualitative research data based on XML schema based on METS, while incorporating existing schema. These include: the Text Encoding Initiative (TEI), the Data Documentation Initiative (DDI) and Dublin Core and a system for identifying segments based on stand off annotation.

Collection of precise address information at the roadside is a difficult task. The challenge is to get the data at the appropriate level of accuracy when it is collected whilst minimising the impact upon the time taken to complete the survey. Using our experience from previous surveys, we have developed a lightweight "portable" version of the TARA Geocoding application, allowing deployment on a handheld device. This allows the surveyor to collect accurate location data electronically and precisely at source. This reduces the amount of post-survey processing of data required. We are currently trialling this software on a small travel-based survey and the results will be presented to the conference. The paper will review the benefits and challenges of hand-held devices versus traditional survey techniques and discuss the implications for designing surveys where location data is sought.

This paper reports on the return rates achieved in a number of internal surveys performed at Bath Spa University over the last twelve years and discusses factors that may have influenced these rates.

The paper will report on the time-scale for returns to e-mail and web-based surveys, suggesting that the majority of responses will be received within a period of seven days. The effect of reminders on the overall return rate will also be examined.

The view for web based surveys on reminders has always been the same as for paper ones; namely that there should be at least a week between each of the two reminders and that the survey closure should be another week after that. Examination of the pattern of responses to a number of previous web based surveys indicated that the majority of responses are within 36 hours of either the initial invitation or a reminder, indicating that people will respond immediately on receipt of the E-mail or not at all.

Initial analysis of the response rates of the four surveys conducted in this way shows that there is no deterioration between the frequent and infrequent pattern of reminders and in fact there is a slight increase in response rates from the frequent groups. The effect of possible other confounding factors such as study year will be considered.

This paper reviews the emerging research record as both panel companies and research agencies begin to look more carefully at the behavior of online panelists. It presents data from individual case studies to show the degree to which online survey results can be influenced by these panelist behaviors. Finally, it describes the steps that researchers can take to ensure data quality when doing research that relies on online access panels.

The Dutch research organization SCP has been trying to improve survey quality and enhance response rates over the past years. This has resulted in truly random samples and good (or at least much better) response rates. However, nonresponse studies indicate that many of the 'additional' respondents belong to groups that were over-represented anyway. As a consequence, the final sample may be even more imbalanced than in a survey with a lower response rate, and final nonrespondents may still differ substantially from late contacts and converted refusers. The presentation will give a short overview of three surveys, describe the extra field efforts and the resulting response rates, and discuss the potential effects on bias. Rather than spending funds on increased fieldwork efforts to indiscriminately haul in additional respondents - it might be preferable to focus on target respondents who are unlikely to participate, and to collect additional information on final nonrespondents.

Decreasing participation in traditional surveys leads to higher costs and less accurate results. Online panels are promoted as the solution to these problems. They are inexpensive as large samples can easily be drawn from databases of many willing respondents.

The paradigm of probability sampling has shown to work well in survey research. Often the impression is created that it also applies to online panel research. Unfortunately, this is not always true. Under-coverage and self-selection may seriously limit the value of their results.

It is sometimes claimed the problems can be reduced by applying weighting adjustment. This is too optimistic. Will discuss the quality of online panels and will present examples of possibly useful applications in academic and governmental research.

Market research considers online panels to be the future. However, there are signs that despite their increasing market share, the rising star of online panels is already on the wane.

The analysis of survey data, collected on a set of response variables defined over a finite population, may benefit from a bird's eye view of their inter-relationships and in particular, of their strengths. This overall analysis should highlight those variables that strongly modify the conditional distribution of another variable, and by contrast, should indicate those which have little affect. We introduce a weighted graph based on measures of independence strength calculated from the population that fulfills this purpose. We show that the graph may be properly defined in terms of population measures without any appeal to super populations, probability modelling or to likelihood. A sample of young women and their smoking behaviours, taken from the General Household Survey is used as an illustration.

If the concept of bounded rationality is accepted then there is a clear need to attempt to record the complex and often non-rational behaviour of decision makers. To address this challenge there has been renewed interest in the application of 'mixed methods' to evaluate socio-economic policies. However, it is widely recognised that the normal 'mixed method' approaches - using standard quantitative survey techniques supported by qualitative methods such as semi-structured interviews - often fail to measure or investigate issues 'outside of the fence'. The aim of this paper is to consider whether these challenges can, in part, be addressed by including projective techniques within the 'mixed methods' approach.

Contextual factors and significant events may influence individual reactions to questions that relate to social and cultural issues. For instance, electoral systems, political scandals and international terrorism may have an impact on trust in politics. One drawback of surveys is that the data they generate from respondents are often isolated from their context and thus potentially misleading (or at least uninformative) about material facts. As a result, individual behaviour and attitudes are described and analysed as if the individuals concerned were living in a society without history, geography, social relationships, social and institutional structures and social and political events. This could be an important source of bias in cross-national surveys. From the start of the European Social Survey (ESS), the need was recognised for a context and event data inventory to provide the necessary background information when analysing survey data. This paper presents an overview of what has been done so far - with examples from different rounds of the ESS - and of the plans for the future.

The assumption for surveys is that respondent behavior or answers to some questions are influenced by significant societal events in various areas. When conducting a survey, the impact of an event must be considered and, whenever possible, controlled especially in survey projects in which different countries participate. However, manual identification of significant events that occur during the data collection phase is a very errorprone and time-consuming task. Therefore, we have developed a procedure to identify events using a combination of two different approaches of the (quantitative) computer-assisted content analysis: the reference text technique and the statistical association approach. On the basis of distinctive features of word usage in a so called reference text corpus and in newspaper texts of a specific time period in which events should be located, words are selected and classified by means of an exploratory factor analysis. The identified factors used as indicators for societal events. Our procedure offers a new opportunity for research in survey-based practice.

In recent years the Social and Cultural Planning Office (SCP) of the Netherlands sponsored two large surveys among ethnic minorities (LAS in 2004 and SIM in 2006). The main goal of the LAS survey was to measure the living conditions of the citizens of ethnic origin in the 50 largest cities of the Netherlands. The SIM Survey was aimed at measuring the attitude of ethnic minorities towards Dutch society and their opinions on integration of ethnic minorities and cultural differences.

In the design and the data collection phase of both surveys we encountered (unexpected) methodological issues that could influence the data quality.

The paper will cover these issues (and trade-offs) and will give some useful suggestions about possible strategies to increase response rates and which techniques to avoid. Also we will briefly discuss which areas to give extra attention to when conducting this type of social research.

The use of computer interviewing programs has contributed to a significant increase in the complexity of survey instruments. Research has shown that survey data can sometimes be affected by related matters arising during interviews. Current methods of documenting the interviewing programs show a lack of standards or conventions, making it difficult for secondary analysts to identify potential interview context effects. The solutions proposed in recent ambitious documentation projects have yet to be implemented widely. In the meantime this paper suggests some simple presentation conventions to be applied whenever a text version of a questionnaire is being prepared, in the hope that these might enable more readily understandable documentation to be published.

This paper looks at the differing profile of respondents who proactively choose to complete surveys on-line when presented with different mode options for completion. Through examining data from the Youth Cohort Study (YCS) 11th cohort, which in all four data collection phases has offered respondents the option of completing the survey questions on-line as an alternative to the .standard. postal questionnaire, we present information about the profile of web responders; examine how they differ from other responders; look at the loyalty of cohort members to a particular mode across different data collection phases and examine at what the findings mean in relation to non-response bias within the YCS series.

This paper describes an implementation of lexical patterns, developed for use with Snap survey software, and designed to make electronic interviews more amenable to respondents whilst providing researchers with in-depth response-interrogation tools. The paper is of interest to all those conducting online and offline research using direct respondent input into electronic questionnaires, and those writing software for such applications. It follows on from other work we have done in this field and draws on work from the fields of computer science and language design.

What is the real state of wireless technology, and what effect will it have on survey research over the coming years? Blending their experience in both survey research and wireless technology, Mark and Aaron will present their view of a wireless future. Not only will wireless technology affect traditionally mobile methods such as face-to-face interviewing and diary studies; it will completely transform how people interact with each other, and how businesses communicate with their customers. And lest we get too comfortable with the current state of wireless innovation - a complex network of cell phones, e-mail, SMS and web browsing - the authors will follow the convergent path of these technologies to show a glimpse of the future. This upbeat, interactive and sometimes controversial presentation will make many people question their core assumptions about the future of survey research.

Time series estimates are derived and produced by the Office for National Statistics (ONS). The ONS has recently developed a software system which enables time series functions to be applied consistently across a wide range of time series. The system uses the X-12-ARIMA software package to carry out a number of time series functions including seasonal adjustment, forecasting and interpolation. This paper gives an overview of the software system and highlights the benefits of an integrated approach that allows time series estimates to be derived in a consistent, timely and transparent way.

Creating efficiencies in Yahoo!'s research operations is highly valued for generating the maximum number of usable insights from a limited research budget. Raising organizational competency for running in-house ("on the network") surveys means use of costly full-service vendors can be reserved for instances where they add the most value (conducting research "off network"). This case study will give insight on the end-client view of the changing MR industry; mainly, the evolution of some end-clients to working directly with more 'links' of the research supply chain (e.g. traditional sub-contractors to the full-service research companies).

Issues of privacy and trust are critical in understanding people's willingness to disclose personal information to web-based services, including to web-based survey systems. In this paper we present the results of two studies that examine the association between privacy, trust and people's disclosure to a web-based survey.

In Study 1, measures of privacy concern were collected, followed six weeks later by a request to complete a web-based survey containing requests for sensitive personal information, alongside measures of trust in the requestor, and perceived privacy related to the specific request. In Study 2, privacy and trust were experimentally manipulated, and disclosure to a web-based survey measured.

As predicted, there was an interaction between privacy and trust, such that the lowest level of disclosure was when low trust was combined with low privacy. However, disclosure was unaffected when there was either high privacy or high trust combined with low trust or privacy.

Two aspects of quality in survey statistics are their timeliness and relevance. A considerable amount of discussion has taken place in the European Union over the timeliness of economic statistics in comparison with the USA. Part of this difference stems from the different approaches to the definition of the period being measured, and the difference of approach has an impact, often hidden, on the relevance (the degree to which the survey measures the required concept). In this paper we review the different ways in which periods are defined in surveys and their properties.

In this paper we will assess the use of audio and video in order to convey the content of the question meaning.

The paper reports results from a field experimental study on the impact of audio and video support in Web surveys on data quality. Within a Web survey among university students a standard interactive online questionnaire was used. A random sub-sample answered a version of the questionnaire that consists of not only written questions but also of corresponding audio files reading the questions to the respondent. Also, a questionnaire version providing video segments showing an interviewer reading the questions to the respondent is tested. Data quality is assessed using standard indicators (item non-response, social desirability scale, social presence scale, non-differentiation of responses and others) as well as the time to complete the survey are used.

Results will be discussed in the light of the feasibility of multimedia Web surveys as well as in terms of the requirements of appropriate survey software.

If the presentation of results from web surveys is aided by the construction of graphical images such as bar charts or two dimensional maps, then why shouldn't the interview process be similarly improved by having respondents interact with graphical images as a means to input their responses?

This paper considers the pragmatic aspects surrounding the deployment and use of graphical image-based controls for collecting web survey response data.

The relative merits of graphical image-based web survey ratings are compared against those of traditional web survey scales. In particular we explore whether interactive web graphical scales can be a direct replacement for traditional web surveys or have more specialised application.

As a busy agency, Network Research has long recognised the need for a high quality system for managing and controlling the entire research process. Frustration with commercial offerings (most survey software focuses either on data collection, the analysis of data or a combination of the two) has led us down a path of in-house development.

Over the last 3 years we have built a sophisticated system that has radically improved the way projects are managed and the information flow both internally and externally. We believe that this system is unique within the research industry and provides us with a notable competitive advantage.

This paper aims to outline the system, the advantages it yields and offer our learning to commercial software suppliers on what an agency such as ours needs from a complete survey solution.

This paper addresses the design, technical implementation and field experiences of using a Web-based data management system (DMS) for the GirlTalk? for Teen Moms Study, being conducted in Washington, D.C. The DMS monitors all activities for screening, recruitment, enrolment, in-depth interviews, randomization, interventions, repeated measures, outcomes, adverse events and follow-up activities for each participant over the course of their two-year enrolment in the study. Field staff from multiple sites use the DMS to manage over one hundred possible activities with a complex series of triggers that lead from one activity to the next. The steps taken in designing the DMS, its technical implementation and lessons learned are presented. This Web-based DMS is compared with a PC-based system used in previous studies. Cost considerations are outlined for the Web-based DMS versus the PC-based system in terms of study size, study duration, number of sites, technical infrastructure, and reusability.

To asses the administrative complexity of mobility for non-european students moving throughout Europe on the ERASMUS MUNDUS Network and E-business Centered Computing programme an extensive evaluation study is being undertaken throughout the period of student mobility. This evaluation is undertaken in a mixed mode with students completing a series of on line surveys and participating in group interviews at each location they are studying at.

This paper introduces the problems encountered with face to face interviewing, then discusses the relevant factors contributing to online disinhibition effect and how this can be used to obtain candid feedback. Finally, experiences of using chat tools to perform group interviews are detailed and an analysis of the effectiveness of this methodology is considered.

Are you an author who sees licence terms as a cause of unnecessary legal fees? Or a user who sees them as 'small print' to be ignored in the rush to install and use the latest application? Or do you recognise them as important, but wish that you understood them more and that they could be simpler?

The paper will explain the more detailed aspects of licensing including many of the current pitfalls. The presentation will also include a light-hearted review of some well known software licences.

Peter Wills and Paul Sampson will explain the key issues in licensing software - why it is necessary, what is needed and what should be avoided - and will show how inappropriate some well known commercial terms are.

Licensing can only be effective when it achieves the right blend of legal, technical and commercial elements.

A paper presented by William O'Connor at the ASC conference "Words instead of Numbers" in September 2006 introduced the pioneering new approach to computer assisted qualitative data analysis called 'Framework' which was due for release in 2007. Developed by the Qualitative Research Unit at the National Centre for Social Research in the mid 1980's 'Framework' is a matrix based tool for qualitative data management. By the time of the ASC 2007 International conference the software will have been released in both standalone and network versions.

The presentation will demonstrate some of the most innovative parts of the application and discuss various features of the new software that can assist researchers in producing robust, credible qualitative evidence. In doing so the wider issues relating to quality in qualitative research and how far the use of software packages in the analytical process can support and enhance transparency will be explored.

This talk considers the potential impacts of the use of video conference technology on surveys. As there appears to be no existing research literature that directly explores the use of this form of technology on surveys the presentation draws on studies of the impacts of video communication in other domains , highlighting potentially applicable insights. Key dimensions which will be considered are the likely impacts on the communication processes during an interview, the social interaction between interviewer and interviewee and the satisfaction of both parties with the experience. In addition, a selection of salient features of surveys will be considered to attempt to identify how the introduction of video technology would affect these known phenomena. The paper thus highlights areas where video technology could add value - and outlines potential pitfalls for surveys if this technology were deployed.

Multimodal interfaces, which I have been developing for nearly a decade, are a promising addition to the set of technologies for automating or supporting survey data collection. These recognize users' speech, eye gaze, gesture, writing and touch; they produce speech, graphical displays, and gestures (on screen or embodied in a conversational agent). They allow the user to choose an input modality, enabling users with various disabilities to use the same underlying system, and some systems allow "composite" input, e.g. when asked "How many bedrooms are there in your house?" a respondent might say "these rooms are bedrooms" and circle the relevant rooms on a floor plan, allowing the system to determine the answer.

Multimodal interfaces can improve automatic speech recognition by using gesture and other input to clarify what was said. In principle they could therefore increase the practicality of speech interfaces for survey tasks.

The honesty of respondents has long been a concern in survey research. With the introduction of new technologies in survey interview methodology a new set of questions arise about how these innovations affect respondents' honesty and disclosure. How, for example, do the new information environments engendered by innovative technologies affect whether a respondent will be more disclosive or more deceptive?

A number of common factors emerge that have important effects on disclosure and deception, including anonymity, private vs. public self-awareness, social presence, and topic sensitivity. Some new findings from the nascent literature examining deception in online contexts, however, provide some novel insights into how information environments affect deception production (e.g., recordability, synchronicity, tailorability) and detection. For instance, in our lab, participants appear to be more deceptive in their telephone conversations than in their emails, suggesting that text-based environments may have some advantages over other forms of responding in survey interviews.

The advent of new technologies for animated or conversational agents raises important questions about the future of survey data collection. Depending on how the next generation of self-administered survey interfaces is designed, we could end up merging the benefits of live interviews and self-administration, or end up with the worst features of both.

We are exploring which features of agents - talking head interviewing systems of the future - will improve respondents' performance and satisfaction, and which will hurt, with the goal of determining which features most warrant development for surveys.

In one study, we investigate the impact of agent visual realism (amount of facial movement) and dialogue capability (ability to converse intelligently) on response accuracy for questions from government surveys. In the second, we discuss the impact of agents. visual realism on respondents. honesty in reporting about sensitive topics; this is compared to responses in face-to-face interviews and ACASI.

This paper will set out the University of Southampton's perspective of Marketing Analytics, It will describe how their new MSc in Marketing Analytics aims to integrate research skills, software training and management expertise. It will argue that Marketing Analytics goes beyond pure quantitative measures of marketing performance to incorporate qualitative marketing research and competitor intelligence. In the teaching of Qualitative Marketing Research and Data Analysis, XSight, the qualitative data analysis software from QSR was used for the first time (at Southampton or indeed any University in Europe). The paper will reflect upon the experience of using this software with a very diverse set of international students that are predominantly numerate. Will the outcome mean 'back to the drawing board' in defining the nature and scope of Marketing Analytics?

The need to break down the traditional 'Agencies set the questions' respondents 'answer them' mentality is the first step introducing the Web 2.0 concept into research. Increasing engagement is the secret to raising response rates and developing longer term loyalty. Web 2.0 technology enables the respondents to actively engage with the researcher through the use of discussion forums, chat rooms, community websites and wiki technology. Community members will be encouraged to have more participation in the research process allowing them to keep their own blogs, upload their own pictures and videos and even allow them to design and ask their own survey questionnaires.

The purpose of this paper is to begin identifying some of the technology challenges and share our experiences in exploring the concept of Web 2.0 in the future of online research.

Quanvert has continued its dominance in the survey reporting space despite the release of other products by competing providers. But is its time coming to an end? Can it continue to deliver to the needs of users, or is it about to be superseded? The paper begins with an overview of the Quanvert application, and moves on to discuss a number of significant developments, which have changed the technological landscape and created new opportunities for taking a fresh look at how to address the needs of the market. It is more than just about replacing Quanvert functionality; it is about understanding the business processes around it. The paper will describe what SPSS has learnt . both from customers and competitors . and how it has responded in order to bring a real alternative to Quanvert to the market.

Real-Time Sampling fundamentally differs from existing, conventional online panel recruitment methods. This approach, previously referred to as "The River", is officially known as "Real-Time Sampling,". It works by recruiting individuals in real time from a network of websites.

This methodology is innovative in that it leverages the many potential online survey respondents who are willing to take a single survey, who may not necessarily want to join a market research Panel.

Based on their answers to screening questions respondents are sent to a proprietary routing system. This ensures that respondents are randomly assigned to open surveys.

The actual qualification takes place in the survey to which the Router assigns the potential respondent.

Each is returned to the website from which they were originated, regardless of whether they qualified and participated in a survey or not.

The use of meta-data as description of statistical datasets has been with us for some time, and the value of this approach is demonstrated by the success of meta-data standards, such as triple-s, SPSS Dimensions and DDI. The need to support discovery of re-sources and services over the internet has widened interest in and use of the term meta-data, focussed by initiatives such as the Dublin Core, the (derived) UK Government Meta-data Standard (e-GMS) and ISO-11179.

The influence of formal ideas from IT systems design extends specifications beyond the structure of the information represented to include the formal representation that are needed in order to use the meta-data.

The Opus Project has attempted to go further than recent proposals and represent the statistical models used in the analysis of data, and thus to provide a link back from results based on statistical information to the assumptions and data used.

This paper traces the history of Triple-S, examines some current features and looks at the actual and potential uses beyond the exchange of data between survey software packages.

Triple-S is currently being investigated by proponents of specialised applications such as Business Intelligence and Mystery Shopping.

This paper provides a look at new technology developments which are relevant to survey computing and what they might offer. It is the latest in a series of similar presentations (about every ten years) which have been given at previous conferences.

Hugh Neffendorf is currently working on the impact of new technology on use and dissemination of official statistics, so the crystal ball is not too dusty. However, it seems to be getting larger and cloudier in some areas, as the pace of technology develops, as the old limitations diminish and as the world becomes hungrier for information. It is noted that the advance of hardware and operating systems can now be taken for granted, while the social implications of technology become more prominent. This presentation will offer a view of a survey environment in which most things now appear possible, but not all will deliver their promise or produce reliable information.