1. Research script design method
Design an opening statement to create an atmosphere, dispel the concerns of the respondents, and make them willing to answer the following research questions.
Because different words will have different psychological effects on the respondents, questions that often look similar will cause respondents to make different responses and different answers due to different wording. Therefore, the words used when asking questions must be very careful, so as not to cause misunderstandings, affect the accuracy of the answers, or even cause resentment to the respondents due to improper wording.
In general, the following principles should be kept in mind when designing problems:
1) Avoid general questions
The original purpose of research is to obtain some specific information. If the question is too generalized and generalized, the answer information provided by the respondent will be of little significance. For example, want to ask the customer if they are satisfied with the way a scene is being answered, so ask the following:
Are you satisfied with the answer of the online robot in the invoice scenario?
Such a question is obviously not specific. Since the answer involves two levels of accuracy and reading experience, it should be asked separately:
Are you satisfied with the accuracy of the online robot's answer in the invoice scenario? / Are you satisfied with the experience of the knowledge content form or knowledge interaction form answered by the online robot in the invoice scenario?
Note: It is necessary to indicate what is the form of knowledge content, such as text or animation; what is the form of knowledge interaction, such as one question and one answer or multiple rounds of dialogue.
2) The language of the questionnaire should be oral, in line with people's habits of conversation, avoid written words, and try to reduce the knowledge threshold of conversation, that is, to distinguish between the words of trainers/AI products and the words of customers' understanding . For example, the professional term of multi-turn dialogue cannot be understood by users without the corresponding knowledge background.
3) Ensure the consistency of professional terminology. Since the knowledge reserve of each surveyed object is different , when it comes to professional knowledge concepts, it needs to be explained first to ensure the consistency of understanding between the two parties in the survey. The technical term explanation here does not conflict with colloquialism, and the core is to ensure the consistency of the respondents' understanding.
4) To prevent inducing and suggestive questions , so as not to affect the thinking of the respondents.
The beginning of the script focuses on inspiring the respondents to be free from constraints and to fully express their opinions, so that the unfamiliar distance between the researchers and the respondents is naturally shortened.
However , it is important to note that the open-ended questions initially arranged must be relatively easy to answer , not highly sensitive questions. Otherwise, if the answer is rejected from the beginning, it will be difficult to continue with the subsequent questions. If the basic information and sensitive information of the respondents are unknown, you can try to place such questions in the second half of the questionnaire. If known, you can add them when sorting out the survey information after the survey.
The general structure of the script includes title, description, body, code number, acknowledgements, and research records.
1) Title. Each script has a research topic. The researcher should clearly define a topic at the outset, reflect the research topic, make it clear at a glance, and enhance the interest and sense of responsibility of the respondents.
2) Description. The script should be preceded by a description. This description can be oral or a text description, explaining the purpose and meaning of the survey, the requirements and precautions for answering the questionnaire, and the name of the research unit and the year, month, and date are also filled in below.
The questionnaire description mainly includes the following contents:
The beginning of the questionnaire mainly includes an introduction and a note, which is a description of the script.
The introduction should include the purpose, meaning, main content of the survey, the organizational unit of the survey, the users of the survey results, and confidentiality measures. Its purpose is to arouse the respondents' attention and interest in answering the questionnaire, so that they can give active support and cooperation to the survey.
The introduction is generally placed at the beginning of the questionnaire and should be small rather than large. The beginning of the interview questionnaire is generally very short: the beginning of the self-administered questionnaire can be longer, but it is generally not more than two or three hundred words.
3) Subject. The main body of the research script is the specificity of the research content, which is the core part of the script, including questions and answers. Formally, questions can be divided into open-ended and closed-ended questions. Generally, closed-ended questions are placed at the front of the script, and open-ended questions are placed at the end of the script.
4) Code number. The purpose of setting the code phone number list number is to ensure the uniqueness of research scripts and content output, and to facilitate the management of research institutions.
5) Acknowledgments. Acknowledgments are to express sincere gratitude to the research subjects for their cooperation.
6) Research records. The function of the phone number list survey record is to record the completion of the survey and the problems that need to be reviewed and revised. The researcher needs to record all the contents of the survey process in detail, including the newly discovered problems during the survey process, and file the new problems into the items to be solved. .
In research work, it is generally necessary to prepare effective recording tools, including notebook computers, notebooks, voice recorders, etc.
2. Research script design considerations
The design of the research script relies on scientific guidance on the one hand, and the work experience of the trainers and AI products on the other hand. Therefore, in the process of script design, trainers need to constantly think about, in order to improve their design ability and ensure the achievement of research goals.