The diffusion of information and communication technologies offers new and economic opportunities for conducting surveys. For instance, sampled business can be contacted by email and then directed to a web page for questionnaire administration. The large proportion of firms endowed with an email account and an access the World Wide Web make email or Internet based surveys a realistic option when targeting a firms’ population.. Anyway a serious problem in an Internet survey is the availability of up to date frames with acceptable coverage and auxiliary information needed for discarding non eligible units and implementing complex designs. Another issue is whether web based questionnaire administration leads to higher response rates and better data quality with respect to more traditional modes such as self administered mail questionnaires.

We dealt with these problems in an experimental survey on the firms an Italian province focused on their use of the Internet and e-business in particular. We implemented a stratified survey design using a frame endowed with telephone numbers but no email address. The target population was given by all firms having access to the Internet. Therefore we designed a rigorous survey strategy.

The following survey strategy was proposed: firms were first contacted by phone, asked whether they had an access to Internet and when this was the case they were invited to participate in the survey. Each telephone contacted firm had the choice to receive an email with the web address of the questionnaire, or receiving a paper version of the questionnaire by fax or regular mail. We used email reminders for respondents who decided for the web questionnaire and telephone call backs for those receiving the paper questionnaire.

In this paper we analyze the results of this experimental survey in terms of the choses interviewing mode, response rates and other data quality issues such as item non-response across different modes.