In measurement, one tries to identify and express the dimensions, degree, capacity or quantity of a given variable. This process involves assigning values as determined by a certain set of rules. Research requires good measurement for its results to be trusted. Good measurement should be characterized by validity and reliability properties. These are the two major psychometric properties of a test for good measurement.
Reliability is the consistency of scores obtained from research instruments and apparatus that have been used in an assessment procedure. The four main forms of reliability include interrater reliability, test-retest, internal consistency and equivalent forms. A reliability coefficient is normally used as a measure of reliability. It refers to a type of correlation used to indicate the consistency of relationship. So as to depict a strong consistency the coefficient should be strong and also positive.
Validity
It is the accuracy of results, interpretations and the conclusions that have been made basing on the test scores of a given assessment. It is not the test or the apparatus used that is valid, but the interpretations and inferences made on the basis of the assessment results that are viewed as either being valid or invalid. For validity to hold, it should be examined with reference to the given use for which the assessment is under consideration. It should be based on evidence showing that the construct in question can accurately be inferred from the given measurement operations.
How they affect the quality and generalization of research results
Reliability and validity are central to the core of measurement. They are useful in analyzing the credibility and truthfulness of findings. The core principles of validity and reliability are useful in qualitative research. Various techniques are employed in recording observations with consistence. These include document studies, participation and interviews among others. This is thus stability reliability in practice. In qualitative studies, it is believed that the subject being considered and researchers have a relationship that evolves with time. A wide variety of sources of data are considered, and multiple techniques of measurement are employed in qualitative research. The phenomena in question also interact with various researchers with the implication that different facets of a subject matter are depicted.
Qualitative research focuses on authenticity as the desired outcome as opposed to only one version of outcome. It is more concerned with providing an honest portrayal of societal experiences of the people under study than anything else. There’s adherence to the main principle of validity; achievement of truthfulness by qualitative researchers. In adhering to this therefore, qualitative studies attempt to develop a sort of link between ideas concerning social life and the reality of what actually takes place.
Factors that may threaten the internal and external validity of a study
If one intends to show that there’s a causal relationship between certain dependent and independent variables in a research study, the extraneous variables that are threats to internal validity should be controlled. These threats that affect internal validity include history, maturation, instrumentation, testing, regression artifact, attrition, and selection as well as additive and interactive effects. Threats to external validity include experimental arrangements’ effects, specificity of variables, selection treatment interference, experimenter effects and multiple treatment interference.
Ways in which a research design may be modified to reduce threats to internal and external validity
So as to minimize threats to external validity, a researcher can undertake measures of ensuring that context, setting as well as sample of study are representative of the society to which he or she intends to study. The use of random selection or by using the proximal similarity theory could be essential here. On the other hand, threats to internal validity can be controlled through the elimination of any differential influences that extraneous variables might be having on the independent variables.
Conclusion
In the principles of good measurement, the main ideas of validity and reliability should be considered. In qualitative research, the researcher attempts to measure with consistency the tight link between the empirical world and the abstract ideas.
References
Christensen, Larry B, R Burke Johnson and Lisa Turner. Research Methods, Design, and Analysis. New Jersey: Prentice Hall , 2011.
Creswell, John W. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. 3, illustrated. New York: Sage Publications, 2009.
Marczyk, Geoffrey R, David DeMatteo and David Festinger. Essentials of Research Design and Methodology. Ed. Alan S Kaufman and Nadeen L Kaufman. New Jersey: John Wiley & Sons, 2005.
Trochim, William M.K. "External Validity." 20 October 2006. www.socialresearchmethods.net. 30 April 2012 <http://www.socialresearchmethods.net/kb/external.php>.
