3. Summary of the survey findings: To increase the number of respondents in the
survey and to improve overall response rates, respondents are often provided
with a summary of some of the survey responses. This too is done via the
Internet. Respondents are sent an email that provides them access to a web
site that will contain the survey findings. As with the survey itself, this is a
password-protected site that is accessible for a limited period (1-2 weeks).
The data processing staff performs machine edits and additional cleaning for the
entire data set. Our edit programs act as a verification of the skip instructions and
other data checks that are written into the online program. The edit programs list
any errors by case number, question number and type. These were then resolved by
senior EDP personnel, who inspected the original file and made appropriate correc-
tions. Complete records were kept of all such procedures.
Reliability of Survey Percentages
The results from any survey sample are subject to sampling variation. The magni-
tude of this variation is measurable and is affected both by the number of interviews
involved and by the level of the percentages expressed in the results.
Exhibit A.1 shows the range of sampling variation that applies to percentage
results for this type of survey. The chances are 95 in 100 that the survey results do
not vary, plus or minus, by more than the indicated number of percentage points from
the results that would have been obtained had interviews been conducted with all per-
sons in the universe represented by the sample.
For example, if the response for a sample size of 300 is 30%, then in 95 out of
100 cases the response of the total population would be between 25% and 35%. Note
that survey results based on subgroups of a small size can be subject to large sampling
error.
123
