DigiPoll - Precision Pollsters. Methodology

Telephone Surveys
We only conduct telephone surveys: they are the fastest, most economical, and easiest to control for quality.

Sampling
A major key to accurate surveying is the fieldwork provider’s ability to follow the socio-demographic distribution of a population reliably. The use of the most sophisticated survey software cannot by itself ensure a good representation of a population. The fieldwork provider has to understand the socio-demographic distribution and create the sampling methods to offer a true representation. The system of randomising a sample also plays a crucial role; the better the randomisation, the more accurate the representation provided.
DigiPoll has invested a lot of time and resources into building systems to study the socio-demographic distributions of a particular region. With experience, DigiPoll has nearly perfected this system for the New Zealand and Australian markets. DigiPoll’s method of sample creation enables the survey samples to be randomised across the country, region, or a smaller defined area.
This unique method has helped to achieve the most precise predictions for the 1996, 1999 and 2002 general elections in New Zealand. Our system not only covers listed telephone numbers, but also traveling members of the population, recent arrivals to the country, and individuals with unlisted telephone numbers.
For delivering a balanced sample, DigiPoll strives for the best possible response rate. We ensure this through repeated recall attempts. The more attempts that are made, the better the chances that busy people, travelers and shift workers will be included in the sample.

Questionnaire design and programming
The reliability of research depends greatly on ability to translate the research design into CATI questionnaire. DigiPoll’s extensive experience in the formulation of questions and response scales, together with the branching and skipping feature of our CATI system, enhances reliability and accuracy of response recording, and, as a result, the survey overall. DigiPoll strongly believes a sincere and focused introduction, tailored for each project, helps increase response rate.

Auto dialing and response rate
DigiPoll employs the most advanced CATI system in the country. Dialing is done automatically through a modem, which reduces interviewer error and time spend, and enables frequent call-back attempts to increase the response rate. Our CATI system includes an elaborate call-back manager, which recalls detailed call-status statistics.

Data Entry
Data entry is primarily by mouse-click, which has proven more accurate that typing of numbers or words. The system provides for fast data entry, avoiding delays in registering responses, helping maintain respondent’s interest and minimizing respondent fatigue.
Data compilation is done directly into the system database by the interviewer, avoiding the transcription errors common with manual questionnaires.
.

Training and Supervision
Interviewing staff are trained and then briefed for each job. An operations manual detailing strict procedures for recruitment and training, quality control and client communication is employed. Surveys are conducted in a laboratory with a supervisor who is able to respond to interviewers' inquiries at any time and detect problems which may arise due to hardware, software, or human error. Interviewers are monitored for training and quality control, and verification call-backs are conducted. Statistical analysis of responses is done during the interview process to monitor the demographic balance of the sample, and to revise the sample if necessary. Response rates and quotas are constantly monitored.

Feedback
The client/researcher is invited to brief the interviewing staff, conduct initial test interviews and listen in to actual interviews while the survey is in progress. A debriefing once the study is completed is encouraged to pass on additional information.