they look like an attractive source of data to practitioners and researchers. They are wondering if big data could substitute web surveys to provide information for social and economic decision making. They will not substitute surveys; the message is that the two sources are complementary, but they require the researcher to consider the existing methodological problems. The big data offer a challenging opportunity to revise the role and questions faced from the surveys and to integrate web survey data with other data sources.
KEY TERMS
Blaise:A software package for computer‐assisted interviewing and survey processing developed by Statistics Netherlands.Census:A way of gathering information about a population in which every element in the population has to complete a questionnaire form.Computer‐assisted interviewing (CAI):A form of interviewing in which the questionnaire is not printed on paper. Questions are asked by a computer program.Computer‐assisted personal interviewing (CAPI):A form of face‐to‐face interviewing in which interviewers use a laptop computer to ask the questions and to record the answers.Computer‐assisted self‐administered questionnaires (CASAQ):A form of data collection in which respondents complete the questionnaires on their own computer or device. See also CASI.Computer‐assisted self‐interviewing (CASI):A form of data collection in which respondents complete the questionnaires on their own computer or device. See also CASAQ.Computer‐assisted telephone interviewing (CATI):A form of telephone interviewing in which interviewers use a computer to ask the questions and to record the answers.E‐mail survey:A form of data collection via the Internet in which respondents are sent a questionnaire that is part of the body text of an e‐mail. The questionnaire is returned by e‐mail after answering the questions in the text.Face‐to‐face interviewing:A form of interviewing where interviewers visit the homes of the respondents (or another location convenient for the respondent). Together, the interviewer and the respondent complete the questionnaire.Mail survey:A form of data collection where paper questionnaire forms are sent to the respondents. After completion of the questionnaires, they are returned to the research organization.Mobile web survey:Self‐administered surveys that can be conducted over mobile web‐capable devices. They are similar to web surveys, but they have also unique features, such as administration on small screens and keyboards, different navigation, and reaching respondents in various situations, factors that can affect response processes.Probability sampling:A form of sampling where selection of elements is a random process. Each element must have a positive and known probability of selection.Purposive sampling:A form of non‐probability sampling in which the selection of the sample is based on the judgment of the researcher as to which elements best fit the criteria of the study.Quota sampling:A form of purposive sampling in which elements are selected from the population in such a way that the distribution of some auxiliary variables matches the population distribution of these variables.Random digit dialing (RDD):A form of sample selection for a telephone survey where random telephone numbers are generated by some kind of computer algorithm.Representative method:A methods proposed by Anders Kiaer in 1896 to select a sample from a population in such a way that it forms a “miniature” of the populations.Straw poll:An informal survey conducted to measure a general feeling of a population. Sample selection is such that it usually does not allow concluding about the population as a whole.Survey:A way of gathering information about a population in which only a sample of elements from the population has to complete a questionnaire form.Telephone interviewing:A form of interviewing in which interviewers call selected persons by telephone. If contact is made with the proper person and this person wants to cooperate, the interview is started and conducted over the telephone.Web scraping:Is data scraping; it is used for extracting data from websites. Note that data scraping is a technique in which a computer program extracts data from human‐readable output coming from another program.Web survey:A form of data collection via the Internet in which respondents complete the questionnaires on the World Wide Web. The questionnaire is accessed by means of a link to a web page.
EXERCISES
1 Exercise 1.1 Which of the following options is not an advantage of computer‐assisted interviewing (CAI) as compared with traditional modes of data collection?Data quality is higher due to included checks.The software is in charge of routing through the questionnaire.CAI leads to higher response rates.Data are processed quicker.
2 Exercise 1.2 What is an advantage of an e‐mail survey over a traditional mail survey?Data quality is higher due to included checks.There is less under‐coverage.Response rates are higher.It has better facilities for navigation through the questionnaire.
3 Exercise 1.3 Why were the first surveys on the Internet e‐mail surveys and not web surveys?E‐mail surveys were cheaper.The World Wide Web did not exist yet.E‐mail surveys are more user friendly.E‐mail surveys require less data communication over the Internet.
4 Exercise 1.4 When should the form‐based approach be preferred over the question‐by‐question approach in a web survey?The questionnaire is very long.The questionnaire contains route instructions and edits.All questions fit on one screen.The survey is a business survey.
5 Exercise 1.5 Which of the four features is typically and advantage of web surveys?There is no under‐coverage.The sample size is always large.A survey can be designed and conducted very quickly.Accurate estimates can always be computed.
6 Exercise 1.6 How to avoid the problem of under‐coverage in a general population web survey?Conduct a mixed‐mode survey.Increase the sample size.Conduct a self‐selection web survey.Replace the web survey by an e‐mail survey.
7 Exercise 1.7 Why source integration is an interesting perspective?To optimize information.To conduct mixed‐mode surveys.To totally avoid web surveys.To run only paper surveys.
REFERENCES
1 Barcaroli, G., Bianchi, G., Bruni, R., Nurra, A., Salamone, S., & Scarnò, M. (2016), Machine learning and statistical inference: the case of Istat survey on ICT. Proceeding of the Italian Statistical Society Conference, SIS, Salerno.
2 Barcaroli, G., Nurra, A., Salamone, S., Scannapieco, M., Scarnò, M., & Summa, D. (2015), Internet as a Data Source in the Istat Survey on ICT in Enterprises. Austrian Journal of Statistics, 44, pp. 31–43. doi:org/10.17713/ajs.v44i2.53.
3 Beręsewicz, M., Lehtonen, R., Reis, F., Di Consiglio, L., & Karlberg, M. (2018), An Overview of Methods for Treating Selectivity in Big Data Sources. Publication Office of the European Union, Luxembourg.
4 Bethlehem, J. G. (1987), The Data Editing Research Project of the Netherlands Central Bureau of Statistics. Proceedings of the Third Annual Research Conference of the US Bureau of the Census, U.S. Bureau of the Census, Washington, DC, pp. 194–203.
5 Bethlehem, J. G. (2009), The Rise of Survey Sampling. Discussion Paper 09015, Statistics Netherlands, The Hague/Heerlen, the Netherlands.
6 Bethlehem, J. G., Cobben, F., & Schouten, B. (2011), Handbook on Nonresponse in Household Surveys. Wiley, Hoboken, NJ.
7 Bethlehem, J. G. & Hofman, L. P. M. (2006), Blaise—Alive and Kicking for 20 Years. Proceedings of the 10th International Blaise Users Conference, Arnhem, the Netherlands, pp. 61–86.
8 Bethlehem, J. G. & Stoop, I. A. L. (2007), Online Panels—A Theft of Paradigm? The Challenges of a Changing World. Proceedings of the Fifth International Conference of the Association of Survey Computing, Southampton, U.K., pp. 113–132.
9 Beukenhorst, D. & Wetzels, W. (2009), A Comparison of Two Mixed‐mode Designs of the Dutch Safety Monitor: Mode Effects, Costs, Logistics. Technical paper DMH 206546, Statistics Netherlands, Methodology Department, Heerlen, the Netherlands.
10 Biffignandi,