Mixed-Mode Surveys

Mixed-mode surveys and market research

To have reliable estimates from market research it is necessary to introduce Mixed Mode. Therefore, it is necessary to integrate CATI questionnaires on cellular (mobile) as well as on fixed telephony with the CAWI.

A mixed-mode survey (CATI and CAWI) has the merit of combining the specificities of the CATI mode to those of the CAWI mode. This allows you to maximize the Response Rate and reduce the potential distortion (bias) generated by the use of only one medium. This is widely documented in the literature and in experiences such as Current Employment Statistics (CES), from the Bureau of Labor Statistics (USA).
Today mixed-mode makes use of the whole spectrum of tools made available by communication technologies. CAMI, MAWI fashions are also used, without neglecting social media, up to online panels.

The advantages of Mixed-Mode

In a CATI survey, the distribution of interviews is rather constant over time, with physiological changes due to external events that cannot be considered. An example is the following distribution in a classic CATI survey on population obtained by means of a graphic elaboration of the Cloudresearch software:

mixed-mode graph

On the contrary, a mixed-mode survey presents a temporal distribution as follows:

mixed mode graph 2

As you can see, the peaks correspond to the sending of e-mails and faxes. This illustrates quite well the different organization and response methods of the mixed-mode survey compared to the CATI survey. Consequently, it also indicates the advantages in the combined use of the tools.

The example proposed by the graph refers to a mixed-mode survey of February 2013 with a sample base of 2.700 units. The day on which the survey began, therefore in correspondence with the sending of the 'letter in advance', the number of questionnaires collected was equal to 240. The day on which the closing email was sent (last recall) there was a second peak of 250 respondents. In the first 15 working days there were collected 2.488 questionnaires via CAWI and fax. In the last week the sample objective was reached with a further 362 CATI interviews with which the under-represented strata were closed.

The quantitative objective in a mixed-mode survey is generally exceeded by a few percentage points (in the proposed example it was exceeded by 14%). This happens because the respondents who have already received the invitation continue to fill in the questionnaires via CAWI or send them by fax even when the CATI phase has begun. Our operative choice is to allow the compilation via CAWI and to input the questionnaires received via fax also in the closed layers so that the overcoming of the set objective is the rule.

Want to carry out a survey using the Mixed-Mode method?
Or maybe you want us to do it for you and provide the final data file?