Finest data management techniques. The most developed way to collect primary data in quantitative research is the questionnaire. The beginning of this section is dedicated to him. A postal questionnaire is the subject of a special in-depth study, because it is a very common way to manage questionnaires in management and requires special methods. Then we present other ways of collecting primary quantitative research data, namely observation and experimentation.
Collection using the questionnaire. The questionnaire allows you to directly ask questions to individuals, pre-determining using the qualitative approach the form of answers through the so-called “closed” questions. This is a basic data collection tool that is well suited for quantitative research, as it can process large samples and establish statistical relationships or numerical comparisons.
Finest resolution data. Three main steps focus on data collection using the questionnaire: the initial construction of the questionnaire with the choice of measurement scales, preliminary tests to verify the reliability and reliability of the measuring device and the final administration. For each of these steps, the researcher must follow certain rules if he wants to get the most relevant and usable data. Both French (mainly research on marketing research and research) and American (quantitative research on methodology) books describe in detail the procedures to be followed.
Our choice is to enlighten the researcher on some fundamental issues. Selection of measurement scales. To build a questionnaire for quantitative research means to create a measuring device. Therefore, it is necessary to select the measurement scales that will be used before solving the problems of writing questions and organizing the questionnaire. Various types of measurement scales (nominal, ordinal, interval or proportional) are presented in this book. In addition to the type of scale, the researcher is faced with the choice between using pre-existing scales or creating his own scales.
Writing questions is hard work. It is important to avoid mistakes in the wording and arrangement of questions, as well as in the proposed answers. Here we summarize some fundamental points that must be observed. The questionnaire usually begins with relatively simple and closed questions. More complex, more complex or open-ended questions are preferably grouped at the end of the document. Whenever possible, questions should be followed in a logical order that emphasizes thematic groups and facilitates the transition from one topic to another.
Two common biases are traps that should be avoided when organizing and formulating questions. The halo effect is an association between a series of sequential questions that are too similar. This can happen when someone goes, for example, to a long group of questions that have the same scale for any modality of answers.
To avoid the halo effect, we can make a change in the form of questions or offer an open question. The effect of pollution is the effect of a question on the next question. To counter this bias, you need to carefully monitor the order of issues. Specialized literature on the development of questionnaires provides fairly specific recommendations for each case.
Finally, it is always necessary to indicate the unit of analysis relating to a question or questions, a number of issues (industrial sector, company, product line, departments), and systematically indicate any unit changes. When the questionnaire includes questions on various topics, it is useful to prepare the respondent’s attention by presenting a new topic with a short sentence separating groups of questions.
After developing the first version of the questionnaire, the researcher must perform preliminary testing to check the form of the questions, their planning and check the understanding of the respondents, as well as the relevance of the proposed response methods. Ideally, the questionnaire should be conducted face-to-face with several respondents in order to also collect their non-verbal answers. Then, preliminary testing is recommended, conducted in accordance with the method of introducing the final questionnaire, if someone wants to comply with the conditions of interaction with respondents.
The data collected during the preliminary tests can also measure the internal reliability of the scales based on statistical procedures of factor analysis, the Cronbach alpha index and the composite accuracy index. This phase allows you to clear the lists of objects, preserving only those that actually measure the studied phenomenon. At the end of the pre-test phase, the questionnaire should be relevant, effective and understandable for both the researcher and the respondents.