Computer Assisted Telephone Interview
Computer Assisted Telephone Interview ( CATI ) describes the support of the telephone interview with the help of the computer. This happens in the following technical areas:
Implementation of the questionnaire
The questionnaire , which is mostly available as a text (file), is implemented in a computer program. The interviewers then conduct the interviews on the screen, and the questionnaire program (CATI program) presents question by question on the screen. The interviewer enters the answer. The entire process logic of the questionnaire is implemented by the computer program.
Management of all telephone numbers (addresses) to be called
If a large number of people are to be questioned in a short period of time, a stringent reminder management of the telephone numbers is essential. Called numbers may be busy, the person you are looking for may not be at home, or an interviewee interrupts the interview and asks to continue two days later. All of these events must always be precisely logged and processed so that this telephone number can be called again at the right minute and brought to the screen for an interviewer. In the case of international projects, the time difference must also be taken into account. In CATI systems, this administration takes care of a component of the overall system.
Replacement or operation of the telephone system
The CATI system either implements the necessary telephone technology itself, or at least controls an existing telephone system and receives and logs the response messages from it. The address management (component 2) recognizes that a telephone number has to be called. The telephone component executes the dialing command, establishes the connection with the interview component if necessary and / or returns the result of the dialing attempt to the address management. Is z. B. busy, then this is reported immediately to the address management and this can schedule the next call attempt.
Monitoring of the telephone studio by the supervisors (telephone managers)
The monitor component enables the supervisor (supervisor, also known as the telephone manager) to monitor the entire studio on his screen at any time. Various windows show who is sitting at which screen, where which project is being worked on, and how many interviews have already been carried out per project. The supervisor can connect to an interviewer's seat using a viewing and hearing function and follow the interview in real time in sound and image. If the called party has not given their approval, listening to a telephone interview is usually illegal or even punishable under Section 201 of the German Criminal Code because of an inadmissible interference with the confidentiality of the spoken word.
Project statistics
The key figures of the status of the various projects are prepared in table and chart form. For example, the project manager can prepare statistics on the current address situation on a daily basis, determine the distribution of interview times or calculate key figures for interviewer performance. The project statistics do not deal with the survey data collected in the project. Rather, it only evaluates information about the project execution.
Support software development
This means the programming environment for the questionnaire developer (e.g. editor, debugger, test driver, runtime environment) and the software for processing the recorded data for the purpose of transferring it to tabulation and evaluation programs. To implement questionnaires, special programming languages are usually used that have only been optimized for precisely this application.
Language support
This component helps to translate a questionnaire program so that the natural language parts of the questionnaire are brought into the new target language without affecting the syntax of the questionnaire program. To a certain extent, a new skin (the interview language) is pulled over a program skeleton (the syntax).
Not all CATI systems have all of these components and some also have additional modules. In any case, CATI is more than just a programmed questionnaire.
Spot checks
The determination of random samples can be done with the CATI using the so-called RLD procedure . Random samples from telephone numbers listed purely (in public directories) or expanded to include a cell phone number boost - to enable households without a landline network to participate - are also common. Nevertheless, the representativeness of the sample suffers from the fact that only people with access to a telephone can be interviewed.
Individual evidence
- ^ Rainer Schnell / Paul B. Hill / Elke Esser: Methods of empirical social research . Munich, Vienna 1999. p. 353.