Reliability can be established empirically by means of repeated
trials. In other words, the selected measuring procedure is subjected to a
number of trials, and if the results are the same for each trial,
reliability is established (Kiefer, 1997). This also means that persons
other than the initial researcher can replicate the measuring procedure and
still produce the same results. There is, therefore, no doubt that any
conclusions, theories, or claims emanating from the research are based upon
Although reliability can be established in empirical terms, Kiefer
(1997) warns that inherent difficulties may still arise with this aspect of
measurement. The ideal is to measure exactly the issue at the heart of the
experiment. The difficulty arises when such focus becomes challenging with
the introduction of elements other than those being measured. This
frequently occurs when the study area is situated in a natural or social
setting. When the focus of study and measurement is for example something
as complex as human behavior, it is unlikely that exactly the same result
will be obtained each time. In this case the measuring device is the
researcher's observations. Because the researcher is human, such
observations are subject not only to flaw, but to multiple interpretations.
Thus, although more consistent and more empirical than validity,
reliability still offers its own set of challenges to the measuring
Validity is a still more difficult concept to pin down. Whereas
reliability is concerned with the processes and instruments by which
research is conducted, validity is concerned with the finished research.
The question here is therefore focused more broadly, and concerns the
accuracy with which a study reflects the concept being studied. Validity
therefore aims to establish the success of measurement, while reliability
...