Improving CATI Fieldwork by blending CATI with SMS data collection
This article is the last in the series of discussing how CATI fieldwork, in general, can be enhanced by the usage of SMS. This article is co-authored with my colleague Carsten. In the previous two articles, we discussed various methods used to improve CATI data collection by means of advanced B2B and B2C sampling.
Data collection in the future will converge more towards a mixed mode of data collection. Therefore, a blended sampling approach assists with multi-mode data collection. Thus, we refer to blended data collection as the use of a single sampling frame with multiple sources of data. In this article, we focus on 1-way SMS messages which consist of the respondent receiving a link to a survey.
A. Blended Data Collection
Data blending begins with the gathering of data from different sources. Once these valid sources are located and loaded, the next step is to merge the useful data into a functioning dataset (sampling frame) while excluding the polluted noise of unnecessary data. The sampling frame could be a client database or a national representative (NatRep) mobile phone frame in a population with high penetration of mobile phone numbers.
In this case, one can use an initial wave of survey invitations via SMS. Since the opening rate of SMS is beyond 90% – most of the people that respond to SMS surveys will complete the mobile survey within a few hours. By using the survey link it is possible to track whether respondents have opened or begun the survey. Furthermore, an extra reminder can be sent out via SMS to respondents, while the respondents which have not reacted can be contacted via CATI.
The benefits of this method are evident:
- Cost savings due to part of the interviews being completed online on a mobile device.
- Higher response rate
- Different characteristics of respondents (some prefer online while others telephone) therefore lower bias.
The timelines and a number of reminders can be chosen depending on the type of survey. In the initial phase of a survey, it is recommended to make use of an A/B split in order to reach maximum efficiency. Respondents can be targeted with SMS messages at any specified time.
B. Pre-notification SMS
Another simple but very effective way to decrease cost while at the same time increasing response rate is the usage of pre-notification SMS. With a pre-notification, it is possible to build up trust, inform the respondent of the purpose of the study and set a calling time.
C. SMS Follow-up questions
Some surveys require extra follow-up questions for clarifications or weighting. In many cases finding the right respondent is tedious. Therefore, it is recommended to collect the mobile phone number of a respondent during the main fieldwork. Not only is it possible to follow-up directly with the right respondent but also make use of SMS to collect extra data either through 1-way or 2-way SMS messages.
This procedure reduces the probability of selecting the wrong contact person within an organization, decreases the cost due to less time spent on agents having to find the right respondent or even removing the necessity for a CATI follow-up completely, as part of the data can be collected via SMS.
D. Pre-Screening via SMS
In low incidence rate (IR) surveys, one of the biggest cost factors is pre-screening. An agent in the CATI facility will have to work through the respondents and check for eligibility. With some surveys having an IR below 10% it is possible to imagine not just the extra cost due to the low incidence but also the decreased motivation of the actual agent. By using SMS pre-screening one can make use of a cost-effective method for finding the correct target population. Of course, pre-screening is more feasible for B2C projects. Nevertheless, when using a client set, it is possible to add a pre-screening question. Pre-screening via 1-way or 2-way SMS can be connected to CATI software via an API in order to collect the response directly from the CATI software.
E. Quality Check via SMS
Many CATI fieldwork agencies currently check around 10% of their interviews either via listening-in of call recordings or alternatively via a follow-up call in which multiple questions are asked.
By using either 2-way SMS or 1-way SMS surveys some basic data can be verified automatically thus saving time and cost. During the fieldwork period, mobile phone numbers should be collected (if not already present) and furthermore respondents should be informed about potential follow-up messages.
Researchers have a lot to gain from SMS integration and should move from only using email follow-ups to integrating SMS, as this would provide for multiple modes of interaction with a respondent and lead to a higher response rate. Email reading rate is currently below 1% and will decrease even further while SMS open rates are beyond 90% and read within the first 3 minutes.
Feel free to get in contact with Sample Solutions (firstname.lastname@example.org) or your CATI software provider to discuss how SMS can be integrated within your current CATI fieldwork processes.