Survey Methodology – CATI CAWI CAPI Mixed Mode Survey
All secrets, pros and cons of each methodology.
With Mixed-mode you can mix multiple methodologies in the same survey to maximize time and quality of your data.
For example, you can start data collection with a phone call and then send a link to complete the survey via web.
Or you can invite respondents to join the online survey and then interviewers will interview the respondents who didn’t answer the CAWI invite with a mobile app on their devices. Mixed-mode started as CATI-CAWI only but right now all advanced pro software allow multiple methods (CATI-CAWI- CAPI) in the same survey.
Mixed-mode can be sequential (I start with one method and then move to another) or concurrent (I use multiple methods at the same time).
If you use concurrent Mixed-Mode, you have to be sure that your software is really fit for an all-in-one data collection (one questionnaire, one database and one output file) for a faster data analysis.
How does mixed mode survey work?
To use Mixed-mode at its best you have to purchase a survey software that is able to work with concurrent methodologies.
“All in one” software allow you to create the questionnaire, associate a list of respondent and then export final data once, for all methodologies used.
It’s important to create a questionnaire valid for all methodologies when you use mixed mode. Some survey software allow dynamic texts that vary according to the specific methodology, for example notes for the CATI interviewer.
In any case, we suggest to design a questionnaire that is suitable both for interviewers or respondents compiling.
Survey design characteristics for each methodology.
|Interviewer Presence||Questionnaire characteristics|
|CATI||YES||Long questionnaires, shorter scales (1-5) and broken into steps (e.g., first asking “satisfied/dissatisfied” then asking “would that be very/somewhat,”), more complex logic as you can ask more questions, ability to hide “don’t know” and “refusal” options.|
|CAPI||YES||You can use multimedia, long questionnaires, shorter (1-5) or longer (1-10) scales, facilitate open-ended questions due with assistance from interviewer probing, also not verbal answers can be noted, ability to hide “don’t know” and “refusal” options.|
|CAWI||NO||You can use multimedia, short questionnaires, disable answers editing, direct questions (yes/no, radio button and check box).|
Mixed mode is good to limit costs and increase data quality at the same time.
In this case, you have to consider the cost of the software, of the devices (if you use CAPI), of the interviewers and the emails. The list of items to consider might seem longer but you have to remember that the use of each single methodology will be much more limited than using just one, both in terms of timing and quantity.
Mixed mode Pros
- Costs reduction
- Higher data quality
- Shorter timing
- Maximum reach of the population
- Each respondent can be reached with his preferred
- High redemption
- Useful for multicountries surveys (different methodologies
can be more or less effective according to national culture)
Mixed mode Cons
- Designing a mixed mode questionnaire can be difficult if texts and logic change according to the methodology
- Higher costs compared to CAWI methodology only
Applied Mixed-Mode use.
CASE 1: CAWI – CATI
PHASE 1: I have a list of respondents with an email address. I mail survey links. Using pro software – with integrated email sender – I can schedule one or more reminders to respondents who didn’t answer the survey.
PHASE 2: To increase the response rate I activate CATI methodology too after a few days. In this case, telephone interviewers will only contact users who didn’t answer the web survey. Alternatively, you can also use the two methodologies at the same time filtering the list according to specific parameters: for example, I can move to CATI only respondents with a phone number or only those who started and then stopped the interview.
CASE 2: CAWI – CAPI
PHASE 1: I have a list of respondents with an email address. I invite them to answer a CAWI questionnaire. For example, the marketing department of a brand invites all their stores to compile the questionnaire. Using pro software – with integrated email sender – I can schedule one or more reminders to respondents who didn’t answer the survey.
PHASE 2: To increase the response rate I activate CAPI methodology too after a few days. In this case, respondents who didn’t answer via web will be assigned to the CAPI interviewers. Each interviewer will have his own respondents to go to. For example, I assign to each face-to-face interviewer stores of a specific city.
CASE 3: CATI – CAWI
PHASE 1: I have a list of respondents with a phone number. CATI interviewers will only administer a first part of the interview asking for personal information, including the email address. This way the respondents become familiar with the survey and are alerted that they’ll receive an email invite soon.
PHASE 2: I schedule an email just to respondents who provided their account. The interview will be resumed from the last question answered by the telephone interviewer.
PHASE 3: In this case, I can plan a third phase when I go back to telephone methodology. After a few days from the email invite, I can contact by phone the respondents who didn’t complete the survey via web.
CASE 4: CAWI – CATI – CAPI
PHASE 1: I have to administer the survey to a list of doctors and I have both the email and the phone number. To limit costs I start my research sending an email invite to complete the questionnaire via web.
PHASE 2: To increase response rate, I contact the doctors to complete the interview by phone. Direct contact with the operator will encourage respondents to answer the survey.
PHASE 3: In this situation, it might be difficult to contact doctors by phone too so I decide to move to a third and last phase: CAPI methodology. A part of the respondents list will be assigned to each interviewer so that they can go and administer the interview face-to-face even with no internet access.