Web Data Collection Research Articles

The Impact of Mail, Web, and Mixed-Mode Data Collection on Participation in Establishment Surveys

The Impact of Mail, Web, and Mixed-Mode Data Collection on Participation in Establishment Surveys

When to Text? How the Timing of Text Message Contacts in Mixed-Mode Surveys Impacts Response

Abstract Using multiple modes of contact has been found to increase survey participation over a single contact mode. Text messaging has emerged as a new mode to contact survey participants in mixed-mode survey designs, especially for surveys that include web and/or phone data collection. However, it is unclear how to best combine text messages with mailings and other outreach contacts to improve response rates and data quality. To explore the effectiveness of using text messaging as a contact mode, we conducted a full factorial experiment that varies the sequencing of text messages with mailing contacts (early versus late reminder) and the time text messages were sent (morning versus afternoon). The experiment was implemented in a follow-up wave of a mixed-mode nationally representative longitudinal survey with two sample groups (Cooperative versus Other Respondents). For Cooperative Respondents, text reminders seemed to be effective at increasing completion rates, with the early text reminder being somewhat more effective than the late text reminder, at least early in the field period. For Other Respondents, text invitations were effective at improving the completion rate, but effects diminished quickly once the invitation letter was sent. Additionally, the early text reminder appears to be more effective than the late text reminder at increasing completion rates for Other Respondents. The sequencing of text messages did not affect data quality across sample groups or substantially impact nonresponse. The time of day the text messages were sent did not affect any of the outcome measures examined.

Unleashing the power of AI: a systematic review of cutting-edge techniques in AI-enhanced scientometrics, webometrics and bibliometrics

PurposeThe study aims to analyze the synergy of artificial intelligence (AI), with scientometrics, webometrics and bibliometrics to unlock and to emphasize the potential of the applications and benefits of AI algorithms in these fields.Design/methodology/approachBy conducting a systematic literature review, our aim is to explore the potential of AI in revolutionizing the methods used to measure and analyze scholarly communication, identify emerging research trends and evaluate the impact of scientific publications. To achieve this, we implemented a comprehensive search strategy across reputable databases such as ProQuest, IEEE Explore, EBSCO, Web of Science and Scopus. Our search encompassed articles published from January 1, 2000, to September 2022, resulting in a thorough review of 61 relevant articles.Findings(1) Regarding scientometrics, the application of AI yields various distinct advantages, such as conducting analyses of publications, citations, research impact prediction, collaboration, research trend analysis and knowledge mapping, in a more objective and reliable framework. (2) In terms of webometrics, AI algorithms are able to enhance web crawling and data collection, web link analysis, web content analysis, social media analysis, web impact analysis and recommender systems. (3) Moreover, automation of data collection, analysis of citations, disambiguation of authors, analysis of co-authorship networks, assessment of research impact, text mining and recommender systems are considered as the potential of AI integration in the field of bibliometrics.Originality/valueThis study covers the particularly new benefits and potential of AI-enhanced scientometrics, webometrics and bibliometrics to highlight the significant prospects of the synergy of this integration through AI.

The Future of Web

The future of web is a dynamic and ever-evolving landscape that continues to shape how information is discovered, analysed, and shared. This research paper presents a comprehensive exploration of the emerging trends and potential implications that will drive the evolution of web research in the coming years. Advances in web crawling and data collection technologies are opening new possibilities for researchers to access and analyse vast amounts of information from an increasingly diverse range of online sources. Coupled with the integration of artificial intelligence and natural language processing, web research is poised to become more efficient and insightful, enabling researchers to extract valuable insights from unstructured data. The emergence of the Semantic Web introduces standardized formats such as RDF and OWL, enhancing data interoperability and allowing for more precise searches and cross-disciplinary connections. However, this transformation also raises ethical concerns surrounding user privacy, data ownership, and the responsible use of AI-driven decision-making tools, which necessitates careful consideration by researchers and policymakers. As web research continues to influence knowledge dissemination, open access initiatives, preprint repositories, and collaborative platforms are reshaping the traditional publishing model. Real-time sharing of research findings becomes increasingly common, necessitating new norms of collaboration and knowledge exchange among the academic community. Moreover, the integration of virtual and augmented reality technologies in web research offers exciting opportunities for data visualization, making complex datasets more accessible and engaging. Virtual reality environments may become common spaces for interdisciplinary collaboration, enabling researchers from diverse fields to collaborate seamlessly. However, along with the prospects, challenges lie ahead. Ensuring data privacy and security in an interconnected web environment will be crucial. Additionally, developing guidelines for responsible AI research and addressing biases in AI algorithms will be essential for maintaining ethical practices in web research

Deep auto‐encoder based clustering algorithm for graph‐based web page recommendation system

SummaryWeb‐page recommendation system (RS) plays a major role in smart web systems. The core technology is the RS, which helps users find their interests or discover item preferences. With the rapid proliferation of information technology, the present era witnessed a remarkable growth in the collection and generation of web data. Every day, it is difficult to project the appropriate information to the user and adds severe complexity to the decision making procedure. Moreover, the existing RS techniques have limitations like huge computation time, high computational cost, over specialization, and sparsity problems. This work presents an effective graph‐based web page recommendation model to overcome these difficulties. Initially, the input data is preprocessed utilizing Natural Language Tool Kit executed in Python. The snapshot‐based dynamic graph model obtains the user‐web page relationship. Weight calculation of user‐item and the user‐user edge is also processed with this graph model. The user‐user edge similarity is used to obtain the user preference vector. Then, unsupervised‐deep auto encoder based density‐based clustering algorithm is proposed to cluster the user preference vector, which groups the visited web page details into k clusters. The web pages are recommended based on the highest score calculated by the page ranking method. Finally, the topmost visited web pages are recommended to web users based on the ranking score. “All the news”, MSNBC, and Weblog datasets are taken for evaluation. The proposed method is evaluated with respect to standard measures like F1 measure, accuracy, recall, precision, and root mean square error. Compared with other existing methods, the proposed technique achieved the highest accuracy, 97.99% for All the news datasets, 93.03% for the MSNBC database, and 90.07% for the Weblog database.

Python Data Analysis and Attribute Information Extraction Method Based on Intelligent Decision System

In order to improve the effect or characteristics of Python data analysis and attribute information extraction, a method for intelligent decision system is proposed. The content of the method is as follows: to create a big data mining model for optimal decision making, to use smart data integration method to integrate data functions, to reconstruct management data evaluation information, and to extract and use management multidimensional information parameters regularly. Characterization methods multidimensional information breakdown and optimization of features are performed, characteristics are classified according to their differences, and management and decision-making optimizations are implemented. Based on linguistic Python, combined with rich and powerful libraries such as regular expression, urllib2, and Beautiful Soup, this paper discusses the methods of building modular web data collection, HTML parsing, and capturing link data. The experimental results show that there is a certain gap in the decision support time of the three methods with the change of iteration times. Among them, the decision support time of the method in this paper is always less than 2 s, while the time of the other two methods is longer. Compared with the other two methods, the decision support time of the method in this paper is shortened by about 1.7 s and 3.1 s, respectively. This is because this method classifies the data attribute gap in decision support, which saves the time of decision support. It is verified that the method in this paper can carry out decision support quickly and has certain reliability. It is proved that intelligent decision system can effectively improve Python data analysis and attribute information extraction.

Counteracting Dark Web Text-Based CAPTCHA with Generative Adversarial Learning for Proactive Cyber Threat Intelligence

Automated monitoring of dark web (DW) platforms on a large scale is the first step toward developing proactive Cyber Threat Intelligence (CTI). While there are efficient methods for collecting data from the surface web, large-scale dark web data collection is often hindered by anti-crawling measures. In particular, text-based CAPTCHA serves as the most prevalent and prohibiting type of these measures in the dark web. Text-based CAPTCHA identifies and blocks automated crawlers by forcing the user to enter a combination of hard-to-recognize alphanumeric characters. In the dark web, CAPTCHA images are meticulously designed with additional background noise and variable character length to prevent automated CAPTCHA breaking. Existing automated CAPTCHA breaking methods have difficulties in overcoming these dark web challenges. As such, solving dark web text-based CAPTCHA has been relying heavily on human involvement, which is labor-intensive and time-consuming. In this study, we propose a novel framework for automated breaking of dark web CAPTCHA to facilitate dark web data collection. This framework encompasses a novel generative method to recognize dark web text-based CAPTCHA with noisy background and variable character length. To eliminate the need for human involvement, the proposed framework utilizes Generative Adversarial Network (GAN) to counteract dark web background noise and leverages an enhanced character segmentation algorithm to handle CAPTCHA images with variable character length. Our proposed framework, DW-GAN, was systematically evaluated on multiple dark web CAPTCHA testbeds. DW-GAN significantly outperformed the state-of-the-art benchmark methods on all datasets, achieving over 94.4% success rate on a carefully collected real-world dark web dataset. We further conducted a case study on an emergent Dark Net Marketplace (DNM) to demonstrate that DW-GAN eliminated human involvement by automatically solving CAPTCHA challenges with no more than three attempts. Our research enables the CTI community to develop advanced, large-scale dark web monitoring. We make DW-GAN code available to the community as an open-source tool in GitHub.

Unit Response and Costs in Web Versus Face-To-Face Data Collection: Comparison of Two Cross-sectional Health Surveys.

BackgroundPotential is seen in web data collection for population health surveys due to its combined cost-effectiveness, implementation ease, and increased internet penetration. Nonetheless, web modes may lead to lower and more selective unit response than traditional modes, and this may increase bias in the measured indicators.ObjectiveThis research assesses the unit response and costs of a web study versus face-to-face (F2F) study.MethodsAlongside the Belgian Health Interview Survey by F2F edition 2018 (BHISF2F; net sample used: 3316), a web survey (Belgian Health Interview Survey by Web [BHISWEB]; net sample used: 1010) was organized. Sociodemographic data on invited individuals was obtained from the national register and census linkages. Unit response rates considering the different sampling probabilities of both surveys were calculated. Logistic regression analyses examined the association between mode system and sociodemographic characteristics for unit nonresponse. The costs per completed web questionnaire were compared with the costs for a completed F2F questionnaire.ResultsThe unit response rate is lower in BHISWEB (18.0%) versus BHISF2F (43.1%). A lower response rate was observed for the web survey among all sociodemographic groups, but the difference was higher among people aged 65 years and older (15.4% vs 45.1%), lower educated people (10.9% vs 38.0%), people with a non-Belgian European nationality (11.4% vs 40.7%), people with a non-European nationality (7.2% vs 38.0%), people living alone (12.6% vs 40.5%), and people living in the Brussels-Capital (12.2% vs 41.8%) region. The sociodemographic characteristics associated with nonresponse are not the same in the 2 studies. Having another European (OR 1.60, 95% CI 1.20-2.13) or non-European nationality (OR 2.57, 95% CI 1.79-3.70) compared to a Belgian nationality and living in the Brussels-Capital (OR 1.72, 95% CI 1.41-2.10) or Walloon (OR 1.47, 95% CI 1.15-1.87) regions compared to the Flemish region are associated with a higher nonresponse only in the BHISWEB study. In BHISF2F, younger people (OR 1.31, 95% CI 1.11-1.54) are more likely to be nonrespondents than older people, and this was not the case in BHISWEB. In both studies, lower educated people have a higher probability of being nonrespondent, but this effect is more pronounced in BHISWEB (low vs high education level: Web, OR 2.71, 95% CI 2.21-3.39 and F2F OR 1.70, 95% CI 1.48-1.95). The BHISWEB study had a considerable advantage; the cost per completed questionnaire was almost 3 times lower (€41 [US $48]) compared with F2F data collection (€111 [US $131]).ConclusionsThe F2F unit response rate was generally higher, yet for certain groups the difference between web and F2F was more limited. Web data collection has a considerable cost advantage. It is therefore worth experimenting with adaptive mixed-mode designs to optimize financial resources without increasing selection bias (eg, only inviting sociodemographic groups who are keener to participate online for web surveys while continuing to focus on increasing F2F response rates for other groups).

Transitioning a Panel Survey from in-person to Predominantly Web Data Collection: Results and Lessons Learned

Abstract Over the last two decades, in-person interviewing costs continued to increase while the data quality advantages traditionally identified with this data collection mode have faded. Consequently, some longitudinal surveys have begun transitioning from in-person to web data collection despite risks to data quality and longitudinal comparability. This paper addresses the major issues involved in the transition process and proposes a multi-sample, multi-phase responsive design that attempts to minimize the data quality risks while preserving the considerable cost savings promised by the transition. The paper describes the design as it was applied to the National Longitudinal Study of Adolescent to Adult Health (Add Health)—a nationally representative panel survey of around 20,000 adolescents selected from grades 7 to 12 (typically 13 to 18 years of age) in the 1994–95 school year. Also described are key results from several experiments embedded within the design and the analysis of mode effects. Also presented are some lessons learned and recommendations for other in-person panel surveys that may be contemplating a similar transition to web or mixed-mode data collection.

Response patterns in the Belgian health interview survey: web versus face-to-face mode

Abstract Background Potential is seen in web data collection for population health surveys due to a combination of its cost-effectiveness and ease of implementation and the increased internet penetration rates. Nonetheless, web modes may lead to lower and more selective unit response rates than traditional modes such as face-to-face (F2F) interviewing and hence may increase bias in the measured indicators. This research assesses the response patterns of a web versus F2F study. Methods A Belgian health interview survey by web (BHISWEB; net sample=1010) was organized parallel to the traditional F2F BHIS 2018 (BHIS2018; net sample=11611; subsample used in this study=2748). Socio-demographic data on invited individuals were obtained from the national register and census linkages. To address the high item-missingness on education level deriving from the census, multiple imputation (m = 20) was applied. Unit response rates considering the different sampling probabilities of both surveys were calculated. Logistic regression analyses examined the impact of mode (web vs. F2F) and interactions between mode and socio-demographic characteristics on unit response. Results The unit response rate was significantly lower in the BHISWEB (18.0% (95% CI: 17.0-19.1)) than in the BHIS2018 (43.1% (95% CI: 41.5-44.7)). A lower web versus F2F response rate was found among all socio-demographic groups, however, this lower web response was less pronounced among youngsters, people cohabiting with household members, higher educated people and native Belgians. Conclusions The F2F unit response rate was generally higher, yet for certain groups the difference between web versus F2F was smaller. It is therefore worthwhile to experiment with adaptive mixed-mode designs to optimize resources without increasing selection bias; e.g. only inviting socio-demographic groups more eager to participate online for web surveys (e.g. youngsters) while remaining to focus on increasing the F2F response rates for other groups. Key messages The advantages of web versus F2F interviewing come against its considerable lower unit response rate. Differences in unit non-response between web and F2F vary between socio-demographic groups.

Moving from Face-to-Face to a Web Panel: Impacts on Measurement Quality

Abstract Time and cost pressures, the availability of alternative sources of data, and societal changes are leading to a move from traditional face-to-face surveys to web or mixed-mode data collection. While we know that there are mode differences between web and face-to-face (presence of an interviewer or not, type of stimuli, etc.), it is not clear to what extent these differences could threaten the comparability of data collected in face-to-face and web surveys. In this article, we investigate the differences in measurement quality between the European Social Survey (ESS) Round 8 and the CROss-National Online Survey (CRONOS) panel. We address three main research questions: (1) Do we observe differences in terms of measurement quality across face-to-face and web for the same people and questions? (2) Can we explain individual-level differences in data quality using respondents’ characteristics? and (3) Does measurement equivalence (metric and scalar) hold across the ESS Round 8 and the CRONOS panel? The results suggest that: (1) in terms of data quality, the measurement mode effect between web and face-to-face as implemented in the ESS (i.e., using show cards) is not very large, (2) none of the variables considered consistently explain individual differences in mode effects, and (3) measurement equivalence often holds for the topics studied.

Forecasts of future scenarios for airport noise based on collection and processing of web data

PurposeThis paper presents an analysis of short-term (2025) scenarios for noise emission from civil air traffic in airport areas.MethodsFlight movements and noise levels at a given airport are predicted using a web-data-informed methodology based on the ECAC Doc.29 model. This methodology, developed by the authors in a previous work, relies on the collection and processing of air traffic web data to reconstruct flight events to be fed into the ECAC model. Three new elements have been included: i) topographic information from digital elevation models, ii) a fleet substitution algorithm to estimate the impact of newer aircraft, and iii) a generator of flight events to simulate the expected traffic increase.ResultsThe effects of these elements are observed in 2025 scenarios for the airports of London Heathrow, Frankfurt and Vienna-Schwechat. The results quantify the noise reduction from new aircraft and its increment due to the air traffic growth forecast by EUROCONTROL.

A Complete Research on Techniques & Technologies of Big Web Data Preparation to Web User Usage Behaviour

The rapid advancements in data digitization, the most powerful inventions of learning methodologies in data collection and reduced cost of data storage further enabled the World Wide Web with immense amount of data at significant rate in all the key domains. The generated web data is non-scalable, high dimensional, widely distributed, heterogeneous, dynamic in nature and having useful insights, and thus, it evolved as big data. This situation creates inevitably increasing opportunities in extracting structured solutions from unstructured weblog data for the present big data researchers. Moreover, to provide value addition to any key domain and derive actionable knowledge for various applications, such as, web usage analysis for improvements in fraud detection, product analysis and customer segmentation, got the focus in big data era by the web analysts. To improve operational performance and to discover hidden insights accurately, a comprehensive process is required to investigate the web user usage behavior by analyzing big web data. Towards this, the authors concentrate on reviewing the techniques and technologies of web data collection and preparation for investigating web user usage behavior effectively. In the present paper, the researchers initially pay an attention to explore web log data preparation methods in the traditional approach. Later, the review emphasizes on Hadoop approach for big data preparation and processing. This approach able to concentrate comprehensively on both the stages: distributed data storage and parallel processing of weblog data and to leverage the strengths of techniques and technologies of individual stages. Moreover, the authors deliberately review the possible potential research paths that results in an improved methodologies for data storage and optimized processing speed in the era of big web data

Ecogeographical rules and the macroecology of food webs

AbstractAimHow do factors such as space, time, climate and other ecological drivers influence food web structure and dynamics? Collections of well‐studied food webs and replicate food webs from the same system that span biogeographical and ecological gradients now enable detailed, quantitative investigation of such questions and help integrate food web ecology and macroecology. Here, we integrate macroecology and food web ecology by focusing on how ecogeographical rules [the latitudinal diversity gradient (LDG), Bergmann's rule, the island rule and Rapoport's rule] are associated with the architecture of food webs.LocationGlobal.Time periodCurrent.Major taxa studiedAll taxa.MethodsWe discuss the implications of each ecogeographical rule for food webs, present predictions for how food web structure will vary with each rule, assess empirical support where available, and discuss how food webs may influence ecogeographical rules. Finally, we recommend systems and approaches for further advancing this research agenda.ResultsWe derived testable predictions for some ecogeographical rules (e.g. LDG, Rapoport's rule), while for others (e.g., Bergmann's and island rules) it is less clear how we would expect food webs to change over macroecological scales. Based on the LDG, we found weak support for both positive and negative relationships between food chain length and latitude and for increased generality and linkage density at higher latitudes. Based on Rapoport's rule, we found support for the prediction that species turnover in food webs is inversely related to latitude.Main conclusionsThe macroecology of food webs goes beyond traditional approaches to biodiversity at macroecological scales by focusing on trophic interactions among species. The collection of food web data for different types of ecosystems across biogeographical gradients is key to advance this research agenda. Further, considering food web interactions as a selection pressure that drives or disrupts ecogeographical rules has the potential to address both mechanisms of and deviations from these macroecological relationships. For these reasons, further integration of macroecology and food webs will help ecologists better understand the assembly, maintenance and change of ecosystems across space and time.

A principle of designing infographic for visualization representation of tourism social big data

With the development of technology, people have more needs for the visual representation of interface information. Visualization, image and interaction have all become the design directions of user interfaces in recent years. The purpose of this study is to collect the user’s perception of the chart through experiments, explore the visualization of the chart presentation, and establish the “The principle of infographic design”. This study used the Octopus web data collection software to collect travel-related social big data of Facebook fan page and produce six different types of charts (a total of ten charts) by Tableau visual software that used as the experimental chart material. Relevant visual factors were derived through experiments. The experimental results found the chart element size, color, color gradients, data, sorting, and chart title will affect the results of the user to watch the chart. This study suggests the following induction 4 points to establish the presentation of information charts to provide chart designer: (1) to appropriate use of color to show the infographic, (2) marked the data to help users quickly understand the information, (3) sort the size of the element, so that users can have a sequential view, and (4) make the best use of different sizes, to convey the difference in the proportion of users.

Response rate differences between web and alternative data collection methods for public health research: a systematic review of the literature.

To systematically review the literature and compare response rates (RRs) of web surveys to alternative data collection methods in the context of epidemiologic and public health studies. We reviewed the literature using PubMed, LILACS, SciELO, WebSM, and Google Scholar databases. We selected epidemiologic and public health studies that considered the general population and used two parallel data collection methods, being one web-based. RR differences were analyzed using two-sample test of proportions, and pooled using random effects. We investigated agreement using Bland-and-Altman, and correlation using Pearson's coefficient. We selected 19 studies (nine randomized trials). The RR of the web-based data collection was 12.9 percentage points (p.p.) lower (95% CI = - 19.0, - 6.8) than the alternative methods, and 15.7 p.p. lower (95% CI = - 24.2, - 7.3) considering only randomized trials. Monetary incentives did not reduce the RR differences. A strong positive correlation (r = 0.83) between the RRs was observed. Web-based data collection present lower RRs compared to alternative methods. However, it is not recommended to interpret this as a meta-analytical evidence due to the high heterogeneity of the studies.

A new standardized data collection system for interdisciplinary thyroid cancer management: Thyroid COBRA

The big data approach offers a powerful alternative to Evidence-based medicine. This approach could guide cancer management thanks to machine learning application to large-scale data. Aim of the Thyroid CoBRA (Consortium for Brachytherapy Data Analysis) project is to develop a standardized web data collection system, focused on thyroid cancer.The Metabolic Radiotherapy Working Group of Italian Association of Radiation Oncology (AIRO) endorsed the implementation of a consortium directed to thyroid cancer management and data collection. The agreement conditions, the ontology of the collected data and the related software services were defined by a multicentre ad hoc working-group (WG).Six Italian cancer centres were firstly started the project, defined and signed the Thyroid COBRA consortium agreement. Three data set tiers were identified: Registry, Procedures and Research. The COBRA-Storage System (C-SS) appeared to be not time-consuming and to be privacy respecting, as data can be extracted directly from the single centre's storage platforms through a secured connection that ensures reliable encryption of sensible data. Automatic data archiving could be directly performed from Image Hospital Storage System or the Radiotherapy Treatment Planning Systems.The C-SS architecture will allow “Cloud storage way” or “distributed learning” approaches for predictive model definition and further clinical decision support tools development.The development of the Thyroid COBRA data Storage System C-SS through a multicentre consortium approach appeared to be a feasible tool in the setup of complex and privacy saving data sharing system oriented to the management of thyroid cancer and in the near future every cancer type.

Miscegenation of scalable and DEP3K performance evaluation of nosql-cassandra for bigdata applications deployed in cloud

The huge data streams generated in the modern world are difficult to analyse using the traditional tools and techniques like relational mysql databases. The data originates in the form of audio, video, pdf files and is highly unstructured in nature. An alternative approach is to use the emerging nosql databases. However, not all nosql databases provide full extent of security. In this work, DEP3K based interfusion of security and scalability is developed, where the large collection of web data collected through web crawler is securely stored in the nosql database cassandra. DEP3K is the double encryption with personal passphrase based private key, a new methodology where the large collection of web data is securely stored and retrieved from nosql-cassandra. Performance changes are brought into consideration before and after encrypting and decrypting the voluminous data and improvement in efficiency is observed with the proposed model.

UIWGViz: An architecture of user interest-based web graph vizualization

Visualization improves the user experience by presenting a pictorial view of the web information. Users are benefited in navigating and searching using the web graph visualization because the web graph presents the relationships among the web documents. To achieve this goal, the web information is processed through several steps from the web data collection to the final visualization. Researchers propose and develop various frameworks to systematically accomplish the required steps. However, to make the visualization more effective to the users, their interests or preferences need to be incorporated to the web graph generation procedures. Towards this goal, this paper presents an architecture to produce and visualize the user interest-based web graph (UIWG). The presented architecture establishes the procedures to model the user interests. To construct and manage the user interest profile, the architecture analyses both implicit and explicit user feedbacks. The architecture also presents the steps to collect and analyse the web data and finds out the appropriate positions to incorporate the user interests to the UIWG generation procedures. According to the architecture, the user interests are integrated to the filtering to remove irrelevant web documents, to the graph generation to connect the web documents according to the user, and to the clustering to group web documents that are similar to a specific user. The UIWG is visualized through an interface and the architecture describes the analysis of the user interactions to capture the implicit user feedback. The transformation of the web data throughout various steps of the architecture are also discussed in this paper.

Do Web and Telephone Produce the Same Number of Changes and Events in a Panel Survey

Measuring change over time is one of the main purposes of longitudinal surveys. With an increase in the use of web as a mode of data collection it is important to assess whether the web mode differs from other modes with respect to the number of changes and events that are captured. We examine whether telephone and web data collection modes are compara­ble with respect to measuring changes over time or experiencing events. Using experimen­tal data from a two-wave pilot of the Swiss Household Panel, we investigate this question for several variables in the domain of work and family. We find differences for the work-related variables, with web respondents more likely to re­port changes. These differences do not disappear once the socio-demographic composition of the sample is taken into consideration. This suggests that these differences are not driven by observed different characteristics of the respondents who may have self-selected into one or the other mode. Contrary to work-related variables, a termination of a relationship was more common in the telephone group. This shows that one mode does not necessarily measure more change or events than another; it may depend on the variable in question. In addition, the difference in the protocol mattered: a web respondent in a household that participated fully by web sometimes differed from a web respondent in a household that had a household interview by phone. Nonetheless, the telephone group differed more from the various web protocols that the web protocols among themselves. With more household panel surveys introducing web questionnaires in combination with more traditional face-to-face and telephone interviews, this study adds to our understand­ing of the potential consequences of mixing modes with respect to longitudinal data analy­sis.