Our mission is to provide high quality data
It is our mission here at Nikkei Research to provide all of our clients with highly reliable research data. We aim to achieve this through stringent quality control based on our ISO-compliant flow of operations, rigorous personal data management based on our personal data protection management system and strict compliance with industry standards and also ensure that our staffs at every stage of the research process are committed to collecting and compiling high quality data.
A commitment to quality
Designing questionnaire to maximize data quality
Scientific data collection (planning and design)
Data needs to be collected in such a way that ensures that it is valid. Although data can be collected by anyone, it just ends up as “noise” unless it is collected scientifically, making the resulting findings unreliable.
Rather than letting anyone answer questions or just leaving surveys in the hands of the interviewers, we follow predetermined rules (scientific methods) covering areas such as the design of target respondents, sampling, the design of questionnaire and data collection control measures, including training and instructions for interviewers, in order to guarantee that data is usable.
1. Designing questionnaire that accurately reflect objectives
We design questionnaire with analysing the details at the root of the problem or issue (actions, awareness, etc.) and collect data on the precise status of respondents. We always take analysis into consideration and design questionnaire so as to ensure that no necessary questions are omitted.
2. Designing the flow of questions from the respondent’s point of view
We always design the flow of questions from the respondent’s point of view in order to improve the quality of each survey, including minimizing erroneous responses, reducing response time and improving the response rate.
In addition to adequately indicating the conditions and subject of each question, out of consideration for the parameters of the response (when, who, what, etc.), we also take care to spare the respondent any unnecessary confusion by ensuring that questions flow in the same way as the respondent’s thought process, from awareness to action and from general statements to specifics.
3. Carefully reviewing wording to eliminate noise and erroneous responses
We follow the principle of using simple expressions on our questionnaire and use precise wording so as to eliminate any ambiguity that can frequently result from specialist terminology or foreign terms (written in katakana in Japanese).
We analyze the meaning, contents and structure of all items so as to ensure that the choices are compiled without any omissions or overlapping meanings.
4. Designing layouts to make it easier to respond to surveys
We employ a range of techniques designed to simplify questions and items so as to make it easier for the respondent to understand each question, including keeping sections that the respondent has to read as short as possible and using underlining or bold text to highlight conditions and other points that need to be made clear.
We also select formats designed to facilitate responses, such as using a matrix if there is a sequence of similar questions for example.
5. Conducting pre-tests in order to adapt surveys to response behavior
We conduct pre-tests when we conduct mail, placement, online and other surveys that respondents have to complete by themselves. This enables us to identify issues in relation to responses, such as any bias on the part of the researcher or the need for further explanations, and improve the relevant questionnaire.
In the case of face-to-face interview, we organize discussion sessions with interviewers in an effort to produce questionnaire that will not pose any problems during the actual interview.
6. Reappraising completed questionnaire in order to refine the question structure
After fieldwork has been completed, we review the completed questionnaire, carefully inspect any problems areas in relation to responses and summarize our findings for future reference.
We analyze aspects such as the percentage of respondents choosing “Other”, the distribution of missing responses and the number of incomplete questionnaire in an effort to produce questionnaire that are easier for respondents to complete in the future.
Quality control at field work
1. Inspections (post-research auditing)
As part of door-to-door and telephone interview, inspections are carried out after fieldwork has been completed. This is designed to detect any bias, errors or any factually inaccurate responses and to prevent them from being mixed in with the data.
The number of inspections carried out is determined based on the amount needed to verify at least 10% of the collected samples. In the event of any impropriety, all of the questionnaires collected by the relevant interviewer are checked.
The inspection process involves our telephone research team phoning the homes of respondents in order to confirm that they have completed the survey properly.
In addition to ensuring the quality of data, inspections also help us to train our interviewers.
Data processing
1. Accurate data compilation
Questionnaire and other responses on paper need to be converted into digital data so that they can be tabulated. In order to accurately input paper data, we conduct pre-input inspections on questionnaire in an effort to minimize input errors. As part of the input process itself, we carry out verification on all samples and use a programmed input screen in order to eliminate any input errors.
2. Data cleaning and editing
Input errors in survey data are checked using a dedicated program. In addition to eliminating human error, this also enables us to carry out logic checks on the consistency and other aspects of respondents’ answers to questions.
If any errors are found in the data, we “clean” data, meaning that we go back to the questionnaire and make corrections, and edit data to enable it to be tabulated if necessary (data conversion, composition of new variables, etc.).
3. Data processing in line with analytical techniques
Experienced analysis personnel map out the best possible questions (measurement methods) in order to obtain the data required for the relevant analytical techniques. We also process and refine data in accordance with rules at the data editing stage, covering issues such as how to handle missing values (missing responses).
We also convert responses into the correct client's designated format(ASCII, EXCEL, SPSS, SAS, ASSUM, etc.).
4. Use of the latest analytical techniques
We provide output using multivariate analysis and a wide range of other data analysis techniques.
In the case of text data such as open ended questions, we use techniques such as text mining and indexing (keyword extraction).
