ABSTRACT
All empirical research is based on data. These are not necessarily quantitative: they could be qualitative also. Signs and symptoms are qualities yet are “measurements” in the statistical sense. Qualitative data too convert to numerics when summarized for a group.
Where should we get the data for medical research? In a clinical setting, they come from patients’ interviews and examinations, and their laboratory and radiological investigations. These are called primary data and are recorded in a questionnaire, schedule, or proforma such as an Excel sheet. Secondary data are existing records of previous patients and other subjects. These can also be used for meaningful research.
All measurements are easily understood statistically as variables since they do vary from person to person. Thus there are qualitative variables and there are quantitative variables. These could be on a nominal, ordinal, or metric scale. This distinction is important since the method of inference is different for different types of variables. The merits and demerits of various types of measurement will help to decide which ones to use for different kinds of assessment in research. All these issues are discussed in this chapter.
Types of measurements such as different scales and qualitative–quantitative, discrete–continuous, and other types are discussed in the first section. Tools of data collection such as questionnaire, schedule, record, interview, and examination are described in the second section. Quality of data such as reliability and validity is discussed in the third section, and assessment of validity of medical tools such as pilot study and pretesting, sensitivity–specificity, and receiver operating characteristic (ROC) curves are presented in the last section.
