This Variables Essay example is published for educational and informational purposes only. If you need a custom essay or research paper on this topic please use our writing services. EssayEmpire.com offers reliable custom essay writing services that can help you to receive high grades and impress your professors with the quality of each essay or research paper you hand in.
A variable (indicator, item) is a superordinated attribute, characteristic, or finding that exists in at least two distinct subordinated categories (classes, groups, units of measurement, values). Cases (individuals) can differ on the variable concerning the category they belong to. All cases of a sample or population can be allocated to variables. All cases assigned to the same category count as ”the same” (concerning this variable). If two cases are allotted to different categories, they are regarded as being dissimilar. Categories should be mutually exclusive, meaning that any individual can only be allocated to one category. They should also be exhaustive (i.e., each individual should be assignable to one category). If for some reason one does not know what category a case belongs to, this is called missing data or missing value.
In surveys, each question in a questionnaire (usually) can be considered as one variable, each possible answer to a question as a category (value). Using a coding system, the answers given have to be coded (i.e., they have to be transformed into figures in order to make them processible by statistics). Here, the problem of measurement arises (i.e., the verbal responses have to be correctly transferred into numbers without distorting their meaning). Defining a concept in a way that it can be measured is called operationalization.
There are multiple ways of classifying types of variables. First, variables can be classified by the number and discernibility of their categories. Secondly, variables can differ concerning the level of analysis (level of aggregation) of the cases concerned. Researchers can study individuals: (1) variables can describe these individuals’ genuine characteristics; (2) relational variables describe a case’s interrelationship to other cases; (3) contextual variables capture a case’s embeddedness in a collective. Researchers can also examine collectives (aggregates, higher levels of analysis). These collectives can be regarded as individual cases themselves, but they also consist of individuals of lower analysis levels. Global (integral) variables are variables describing characteristics genuine to the aggregate. Analytical (aggregative) variables have to be calculated from variables measuring characteristics of lower-level cases. Structural variables are calculated from information on the relation between lower-level cases. Thirdly, variables differ in level of measurement (scale in the broader sense); that is, on the question how categories can be arranged. The higher the scale type, the more severe are the measurement rules, but the more statistical methods are allowed as well. Fourthly, variables differ on how they can be operationalized. Manifest variables can be observed and measured directly, while latent variables cannot be directly observed but have to be reconstructed using special methods, such as factor analysis. Fifthly, for some variables, individual values can be interpreted directly, without knowing other cases’ values. For other variables, one needs to know the whole range of values to assess the meaning of a single case’s value.
- Bartholomew, D. J. (ed.) (2006) Measurement, 4 vols. Sage, Thousand Oaks, CA.