Home     About Us     Web Panel     Products      Contact Us	עברית

Info Definition Advantages Panel versus phone survey The ethical code

	Panel versus phone surveys
	In a study by Dennis and colleagues in 2005, possible sources for the differences between phone surveys and those performed on Internet panels were examined. The researchers proposed a number of hypotheses for the possible differences, including: the method of data collection used on respondents – random appeals by researchers during telephone surveys as opposed to a panel established in advance, representation of the sample, the means by which data was collected using both methods, the panelists’ experience during previous surveys via the Internet, and the difference between verbal (telephone) and visual (panel) surveys. The results proved that the difference between the two methods derived from the manner in which data was collected (telephonically - verbal, Internet – visual), and from previous experience acquired by panelists. The differences between visual collection and verbal collection found in the research are very similar to those found in similar research between phone (verbal) and mail (visual) surveys (Dilman & Tranai 1992; Dilman & colleagues1996). Likewise with phone and face-to-face surveys (Dilman 1994). In all studies, it was found that phone surveys were characterized by a tendency for the respondents to give more positive answers on the respondent scale, while Internet panel respondents used more diverse answers from the options presented to them, hence making their answers ‘untainted’. Seniority of panelist respondents and experience in answering many surveys could act in either way: On the one hand, studies show that the ‘adaptation’ of panelists to Internet surveys makes it more comfortable, allowing more authentic answers. On the other hand, the panelist may become ‘professional’. The solution is to have a ‘cooling’ period between surveys, as is the case with panel respondents. Other biases in phone surveys may stem from the interviewers’ interaction with respondents, which changes from person to person and may have an unwanted influence on the answers. Likewise the interviewer has control over the pace of the survey and may apply pressure to the respondent (drawing too quickly, monotonous etc). These biases do not exist in Internet panels where panelists answer the questions without any additional human intervention.