Abstract
This chapter describes the positive features of the Internet for survey activity and examines some of the challenges of conducting surveys via the Internet by looking at methodological issues such as coverage, sample selection, non-response and data quality.
TopTypes Of Internet Surveys
The essential idea of an “internet-based” survey is either 1) rather than mailing a paper survey, a respondent is given a hyperlink to a web site containing the survey – web survey - or, 2) a questionnaire is sent to a respondent via e-mail, possibly as an attachment – e-mail survey. However this is only the baseline idea because diversity is the key characteristic of internet-based surveys. Unlike other modes of data collection, where the method tells us something about both the sampling process and the data collection method, the term “internet survey” is too broad to give us much useful information about how the study was conducted.
In an effort to classify the most common types of internet-based surveys, Couper (2000) and Fricker Jr, 2006) have suggested a division based on the type of sampling methods – probabilistic or non-probabilistic - and the most generally used internet-based survey mode – the web or the e-mail.
The main distinguishing feature between probabilistic and non-probabilistic selection in internet surveys is whether or not the individual is left to choose to participate in the survey (“opt-in”). While in probabilistic selection the respondent is selected by a random procedure established by the survey researcher, for non-probabilistic selection, either a convenience sample is drawn or the survey is distributed/advertised in some manner and it is left up to those exposed to the survey to choose to opt in. Table 1 contains the main types of internet-based surveys and is followed by a brief description of each type of survey.
Key Terms in this Chapter
Population: set of units (e.g. individuals, companies, …) with at least one common characteristic and about which one or several others characteristics are to be studied.
Survey: a type of research which consists in studying a population by observing a sample of units from that population.
Non-probabilistic (or Non-Random) Sampling: a procedure to select the sample which does not guarantee that every population’ unit has some probability of being selected. When there is no sampling frame available the sample must be selected under a non-probabilistic sampling design.
Sampling Frame: list of units which represents the whole survey population; when a sampling frame is available it is used to select the sample.
Probabilistic (or Random) Sampling: a procedure to select the sample which guarantees that every population’ unit has some probability of being selected.
Non-Response Rate: percentage of people contacted that did not respond to the survey.