The Computer-Assisted Personal Interview
(CAPI)
Keywords: CAPI; Survey; Questionnaire design; Quality Control.
1. Introduction
Computer-Assisted Personal Interviewing (CAPI) refers to interviews conducted using handheld devices such as tablets. The questionnaire is configured using CAPI software and is loaded onto the device. The interviewer asks the questions to the respondent, and inputs the responses directly into the device. This electronic data capture approach eliminates the need for post-interview data entry and thus allows for immediate data retrieval.
Since 2016 Geostat has started testing the questionnaires using CAPI Software. We have to make certain changes in questionnaires due to CAPI functionality. By the end of 2017 we plan to introduce the pilot version of agriculture survey.
This paper describes the advantage and the quality control for implementation of CAPI at Geostat, using the World Bank Software.
2. The CAPI Advantage
Introducing CAPI at Geostat reduces the time lag between data collection and data analysis. Since manual coding of the responses recorded with pen and the paper is not necessary and data validation is done in time of data collection, the information is ready for statistical analysis as soon as surveying is completed.
In addition, using tablet devices for interviewing gives other benefits:
- Data validation at entry
Surveys can contain validation data that make it impossible to enter values outside a given range. Supervisors may also view and check the collected information as soon as the enumerators finish the interviews, together with possible error reports. Automated routing reduces the incidence of missing data.
- Real-time Update
Changes in the structure of the questionnaire can be instantly reflected on the interviewers' devices. This allows for last-minute updates or error corrections. Unlike to the PAPI, Changes or corrections of the questionnaires have no cost.
- GPS Monitoring
The interview is normally conducted in the respondent's home or, although rarely, in another agreed upon location. GPS field in the questionnaire gives coordinates about respondent’s home. This information is used by monitoring team.
2.1 Interviewer training
Interviewer experience of CAPI and good training were critical elements in a successful transition to CAPI. The interviewers will use Android Tablets to administer the interview and the Head Office will be responsible for providing specific training for interviewers on using the tablets.
3. Interviewing a statistical survey using CAPI software
Preparing to the survey start with compiling the sample frame at the headquarters. At the same time, the survey designers can collaborate on creating the structure of the questionnaire using the visual Questionnaire Designer tool.
As soon as the questionnaire structure has been approved, the sample is divided into parts, each one being assigned to a team of enumerators.
The team leaders plan the data collection process and assign households to individual interviewers.
The enumerators visit the households and conduct the interviews. If they need to replace their tablet, for example because of hardware malfunction or running out of battery power, they can continue work using another one, as the data are synchronised across devices.
The completed assignments are uploaded to the server. The team leader monitors the progress and, for every interview conducted can:
- approve it, so that it is uploaded to the headquarters server
- or reject it, and force the enumerator to conduct it again (or reassign it to a different enumerator)
The approved questionnaires are uploaded to the headquarters servers. The complete dataset can be downloaded and analysed using statistical software.
4. Implications on quality
CAPI has a solely positive influence on the quality of statistical processes and therefore – directly on the quality of statistics. Implementation of CAPI will allow to improve our activities in line with the ES Code of Practice.
Introduction of CAPI and use of the single application in various surveys will help to ensure further improvement in timeliness and punctuality of statistics (Principle 13 of ES CoP), reduction in statistical survey and data processing costs – the same amount of job will be done with less resources (Principle 10), support standardization of statistical processes and ultimately data coherence and comparability improvement (Principle 13). A computerized interview will also have a substantial positive influence on respondent burden – it will enhance the reliability and validity of interviews, thus reducing the need for follow-up visits to the respondent will decrease (Principle 9).
5. Conclusion
This paper has considered the use of CAPI in statistical surveys and sought to shed light on some of the advantages of using this method, particularly for complicated surveys. CAPI effectively and effortlessly prevents error in the data collected through its in-built automated routing and consistency checks. This significantly reduces the need for back-end processing and data cleaning as the data is more consistent and reliable. This in turn leads to cost savings for the organisations in involved and reduces the time between fieldwork and publication of final results. We expect that these benefits will continue to increase in the future as more sophisticated technology becomes available, enabling CAPI to perform an even wider range of survey tasks.
CAPI can results in significant improvements to the quality of the data collected, whilst at the same time reducing the time, effort, and cost associated with quality control activities. Furthermore, CAPI can also be a useful tool in the evaluation of survey quality, enabling managers to perform a range of tests of the overall data and certain enumerators if necessary.
2