Factors considered before collecting data

Answer 1

While many factors will influence the choice of an appropriate data collection method, choosing the most suitable method will increase the likelihood of achieving a higher response rate.

Here are a few guidelines to consider.

1) Target population: Consider conducting a telephone survey when the sampling population is the general public-that is, the national adult population of Canada. Online data collection has presented a few problems for surveys of the entire Canadian adult population, particularly coverage and sample frame issues.2 internet access is widespread but is still not as widespread as telephone access, and Internet use among certain types of audiences is more limited than among others. In addition to these coverage problems, there is no complete sample list3 available for Internet users (see BP 1.1.1). While the Internet is an excellent data collection method for certain audiences and research objectives, at the time of this report (March 2007), it has greater coverage and sampling limitations than some other methods.

2) Sample frame: The completeness of the sample frame is critical to limiting coverage error. Telephone surveys currently provide the best coverage of the general population, but organizations surveying smaller target populations may wish to consider the quality and composition of available e-mail lists. For example, online data collection might be appropriate and advisable when surveying special audiences, such as scientists and academics. Their e-mail addresses are attainable, and it is often difficult to contact such people by telephone because they travel, rely on voicemail or have gatekeepers, for example.

3) Size of budget: Collecting data by telephone typically costs more than using self-administered data collection methods. When the budget is limited, consider using online, mail, fax or e-mail surveys rather than telephone data collection.

4) Length of time available for fieldwork: If data are required very quickly-that is, overnight or within a few days-and the response rate is not an issue, consider using telephone or online data collection. However, use the latter only where coverage issues are not factors in the decision-making.

5) Level of data precision needed: A telephone survey is appropriate when a high degree of accuracy or statistical reliability is required for general public surveys.Currently, there is no method for selecting random samples from general e-mail addresses, which means that probability sampling is not yet possible using online data collection. This concern may not be important if the research sponsor only requires "directional" information-that is, data that provide approximate magnitudes, rather than precision. While there are studies that show Internet panels have produced results similar to those of comparable telephone samples (Berrens et al., 2003), online surveying is a contentious issue among survey researchers. It is evolving quickly, based on new data and learning.

6) Research objectives: Telephone surveys are appropriate for questionnaires shorter than 20 minutes and studies where respondents can respond to questions quickly. Avoid conducting a telephone survey when the questionnaire lasts over 30 minutes; when respondents need to check things, such as financial information in their files; or when interviewers need to use visual aids to explain concepts or complex scales. Under these circumstances, a self-administered survey might be more appropriate.