Probit recruits its flag-ship panel using a random digit dial (RDD) process for sampling from a blended land-line cell-phone (aka full coverage) frame. The distribution of the recruit process is meant to mirror the actual population in Canada (as defined by Statistics Canada). As such our more than 90,000 member panel can be considered representative of the general public in Canada (meaning that the incidence of a given target population within our panel very closely resembles the public at large).
Stage 1 of this process uses Interactive Voice Response (IVR) technology, which allows respondents to enter their question responses by punching the keypad on their phone, rather than telling them to an operator. The recruit contact begins when the automated IVR call is placed to either a landline or a cell-phone line. Once we have the person on the line we introduce the concept of doing surveys with Probit ‘from time to time’.
Stage 2 the process involves our highly trained recruiters contacting those reached by IVR (those who expressed interest in becoming a panel member). During this call, the nature of the panel is explained in greater detail (as are our privacy policies) and demographic information is collected. The on-line/off-line status of the individual is also ascertained in order to determine the type of surveys they will be asked to complete (i.e., on-line by web or offline by telephone or mail). The fact that every member of the panel has a live telephone conversation with one of our recruiters to confirm their administrative data is very rare in Canada today. It is this extra step, however, that gives us absolute confidence in the administrative data we have on our panel, and our panel metrics.
In order to increase the likelihood of participation in future surveys, ongoing contact is maintained with our panel members through a variety of means (e.g., frequent newsletters). Survey invitations, however, are kept to a minimum (i.e., often no more than once a month).
Probit offers clients a whole suite of research modes to choose from including telephone only, on-line only, mail-only, or any mixture of those modes. Having the flexibility to use a cost effective panel that fields in any mode is a particularly compelling proposition for many clients.
Regardless of the mode selected, the survey sampling process is the same. First, the target population is determined (i.e., adult Canadians, recent PSE graduates, etc.) and a population analysis with Census 2006/11 data is conducted. Second, a random sample from the Probit frame (taking into account composite weighting features that balance the sample according to a propensity to respond) is taken. Third, the survey goes to field, where, regardless of mode, multiple contacts are attempted in order to bolster response rates.
Once the data collection process is complete, results are analyzed and, if necessary, rim-weighting is applied to the data tables (and delivered in the data file).