Reliability Essay ⋆ Communication Essay Examples ⋆ EssayEmpire

This Reliability 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.

‘Reliability’ is the ability to rely on something, a tool, a service, or data. In empirical research, unreliable data limit their use as evidence for theories or scientific arguments. Here, reliability is inferred from the degree of agreement observed among data generated by several independent methods.

There are many sources of unreliability. Measuring instruments may malfunction or be affected by spurious causes. Scientific observers may be influenced by the theory they are pursuing. Eyewitnesses’ accounts of events could be biased by selective recall, self-interests, or information obtained subsequent to the event. Coders employed by content analysts may disagree on their reading of given texts.

The use of data faces two epistemological limitations that motivate reliability assessments: (1) when the phenomena of interest are transitory, it is empirically impossible to verify whether the data about these phenomena are accurate. This is so for earthquakes, historical events, and conversations. Some leave traces behind. Most are known through records kept for a variety of often institutional reasons; (2) even persisting phenomena may not be unambiguously analyzable, in which case the act of observing and recording constitutes the phenomena studied. For example, measuring time presupposes a clock. The meanings of texts arise in processes of reading. Attitudes do not exist in someone’s head; they are created by psychological measurements. Relying on data as unquestionable givens may lead to invalid conclusions.

Unable to refer to the phenomena in place of which data are examined, all one can observe is the agreement among several methods of generating data from the same phenomena. Reliability assumes that above chance agreement among methods can happen only if they respond to something real, not what the phenomena are.

Three concepts of reliability can be distinguished: (1) in theories of measurement, reliability provides assurances that the process of generating data is not influenced by circumstances that are extraneous to observation, description, or measurement. For example, when observers or coders generate records, one needs to assure that their coding instructions yield the same data no matter who is employed; (2) interpretation theory acknowledges that researchers may have different backgrounds, interests, and theoretical orientations which can lead them to different interpretations of the same data. Accordingly, interpretations are reliable when, after accounting for explainable discrepancies, three or more interpretations come to the same conclusion about what is (triangulation); (3) in pursuit of psychometric tests, researchers consider a measure to be reliable when it correlates with other not necessarily identical measures across time, individuals, and situation.

Further, three kinds of reliability can be distinguished: stability (test-retest reliability), reproducibility (test-test reliability), and accuracy (test-standard reliability). Accuracy is the strongest reliability measure, but rarely obtainable for the frequent lack of available standards. Data for measuring reproducibility require two or more independently obtained sets of data describing the same set of phenomena. The cut-off point below which the data need to be rejected as too unreliable to precede with their analysis should be determined by the probability and costs of leading to wrong conclusions

Bibliography:

Krippendorff, K. (2013). Content analysis: An introduction to its methodology, 3rd edn. Thousand Oaks, CA: Sage.
Merrigan, G. & Huston, C. L. (2009). Communication research methods, 2nd edn. Oxford: Oxford University Press.