Continental Research
!
4500 Colley Avenue
!
Norfolk, VA 23508
13
Methodology
(cont’d)
As a quality control and to minimize the bias caused by females answering landlines more
frequently, a statistical technique was used to select the adult in the household who would be
asked to participate. Each landline number was designated as “preference female” or
“preference male” prior to dialing the number. Naturally, all survey participants were to be over
the age of 18.
To ensure a representative sample, each phone number selected for inclusion in the study was
called at least six times (on different days) before a substitute phone number was chosen. This
six-attempt method produces a demographically well-balanced sample and ensures that very
busy people are able to be reached. In addition, Zip Code quotas were maintained.
The interviewers usedComputer-Assisted Telephone Interviewing (CATI) technology to read the
survey questions verbatim and key each response directly into the computer. A supervisor, who
was present at all times, electronically monitored the interviewers' work. He listened to each
conversation and visually observed the recording of survey responses using Dameware
software. The supervisor was able to fully monitor over 38% and partially monitor an additional
25% of all interviews. This is far in excess of the 5 - 10% industry standard for validation.
At the close of each phone shift, a short interviewer de-briefing session was held to discuss the
survey's progress and how residents were responding. These meetings provide useful
anecdotal information that is considered when analyzing the tabulated findings.
When the fieldwork was completed, the open-ended items were analyzed, grouped into
categories, and assigned a numeric code to facilitate analysis. Responses to the open-ended
item about the Chesapeake Jubilee were grouped into categories.
To prepare the 323 surveys for analysis, an ASCII data file was read into a lengthy IBM-SPSS
(Statistical Package for the Social Sciences) computer program. The data were then tabulated
and analyzed, and the results are displayed on the following pages. For the reader’s
convenience, the question numbers precede each table. The questionnaire can be found in an
Appendix. Trend data from 1998 thru 2014 can also be found in an Appendix. A set of
supplementary crosstabulations (under separate cover) was also prepared to show more
detailed breakouts of the data.
