Variables, Scales of Measurement, Validity & Reliability

Variables, Scales of Measurement, Validity & Reliability

1. Variables

   • the factors that you are interested
     • any quantitative measurement of an entity, attribute, or phenomenon being studied in the study sample.
   • Independent aka Study/Treatment variable
     • one that is assumed to remain constant or uninfluenced by changing conditions or circumstances in the study.
     • can change the value of the dependent variable (DV)..
     • cause of the DV and should always precede the DV and has a potential influence on it
     • FIXED
     • Values= Levels
   • Dependent aka Outcome variable
     • changes in ways that are determined by the influence of an independent variable or external factors.
       • value of the DV can be affected by the IV
         • value of the DV depends on the action of levels of the IV
     • the phenomenon that the researcher seeks to understand, explain or predict.
   • Confounding
     • sometimes called extraneous or intervening) variable(s)

2. Aim of research

   • identify how independent (or study) variables
     • influence a dependent (or outcome) variable
       • preferably to develop a generalisable model (through statistical analysis)
         • can explain and predict that relationship
           • identification of variable characteristics is very important in planning, conducting and critically examining research.

3. Measurements

   • involves the process of description and quantification in research
     • relates to the recording of physical or behavioural characteristics by assigning a value to aspects
     • quality, quantity, frequency or degree of these attributes.
   • use of an instrument or tool that is capable of measuring the magnitude of the attribute (or variable) under evaluation by using a calibrated scale
   • involves assigning a value (usually numerical) to an object
     • can be measured directly (length, width, death rate) whilst others can only be indirectly measured through an abstraction
       • Categorical
         • variables with discrete values rather than incremental placement along a continuum
         • assign values to categories
           • gender, diagnosis and country of birth
         • measured using nominal or ordinal scales
       • Continuous
         • measured along an infinitely divisible scale.
           • one that can take on a large range of values representing a continuum.
         • height, weight, blood pressure and age
         • measured using interval or ratio scales.
         • Continuous tranformed into categorical but not vice versa