The 1996 Pew Research Center Survey of Technology is a telephone re-interview survey conducted in the continental United States among 1,003 adults, 18 years of age or older, most of whom were identified in previous surveys as online users. Interviews were conducted from October 21 through October 31, 1996. The margin of error for this survey is plus or minus 3 percentage points for the total sample at the 95% level of confidence.

Sample Design
The sample for this survey was comprised of respondents to previous Pew Research Center surveys who had identified themselves as online users, and a small random sample of previously identified computer users who earlier said they did not go online. The surveys sampled included the 1995 Times Mirror Technology survey as well as several Center News Interest Index surveys conducted over the last year.

The sample was released for interviewing in replicates. Using replicates to control the release of sample to the field ensures that the complete call procedures are followed for the entire sample.

At least six attempts were made to complete an interview at every sampled telephone number. The calls were staggered over times of day and days of the week to maximize the chances of making a contact with a respondent. In each contacted household interviewers asked to speak with the person in the household who had been interviewed before, identifying them by their gender and age. Respondents were then screened to determine if they are currently online users. Those who said they no longer went online were only asked one open-ended question and are not included in the final 1,003 interviews represented here.

Weighting
Non-response in telephone interview surveys produces some known biases in survey- derived estimates because participation tends to vary for different subgroups of the population, and these subgroups are likely to also vary on questions of substantive interest. For example, men are more difficult to reach at home by telephone than women are, and people with relatively low educational attainment are less likely than others to agree to participate in telephone surveys.

This survey was weighted in analysis to bring the demographic profile of the final sample of online users into alignment with the demographic profile of all online users pulled from previous surveys. All of the surveys from which respondents were drawn were nationally representative RDD samples of the general population 18 or older, and respondents had been weighted within these national surveys to the most recently available Census data (the 1994 Current Population Survey). The weighting parameters for this survey came from an analysis of all potential respondents (online users and non-online computer users) pulled from the surveys identified above, and represent the weighted demographic profile of the total pool of potential respondents. This weighting strategy produces a sample of online users that is weighted to the best estimate of the demographic profile of such respondents and takes into account the potential non-response bias from not being able to re-contact each person from the original sample.