【明報專訊】There are many different types of data that can be collected by a researcher. It is important that the nature of the data is explained initially, again in the methodology, and then explored thoroughly within the findings section. Data can be observational, experimental and, if based on personal experience and/or research experience, experiential. The data should be thoroughly analysed to look for interesting patterns and areas (exploratory) before these aspects are focused on and communicated to the reader (explanatory). Data may have been extracted from a computer model or simulation (computational) or from the study of people and culture (ethnographic). The dataset may involve just a single variable (univariate), two variables (bivariate), or more than one target variable (multivariate). Normative data, a set that establishes a representative value or "norm", is different from normalising the data, which involves adjusting values so they align or are made clearer. Data may be combined (pooled) or in a separate unit but part of a larger group (nested). Nominal data or data with finite values (discrete) are unlikely to be as objective (quantifiable) as interval data are.