Methodology to correct the unit non-response bias in the case of the KOF ETH Survey 2000 on organization and information technologies

In order to correct the bias due to unit non-response for the KOF ETH Zurich's business (mail) surveys, we usually use the results of a second (phone) survey by the non-respondents. Taking the case of the survey 2000 on "Organization and Information Technologies" in the Swiss economy, we describe how to build the sample of this second survey and how to use the collected data. Actually, we show how to generate new correcting weights to correct the non-response bias. -- Unit non-response bias ; weighting ; calibration ; Information Technologies