Ten Secrets of Successful Students

Validity and Reliability of Engineering Analytical Methods & Processes

Engineering Project Management

The Seven Deadly Sins of Modernity

The Paradox's of Our Time

If I had my life to live again

The world NEEDS engineers

Graduate Student Research Organization Questions

What does "completed student work" mean?

Validity and Reliability of Engineering Analytical Methods & Processes

Engineering analytical methods and processes and changes (suggestions, recommendations) implicated by such processes should be valid and reliable. How can the properties of validity and reliability be determined?

Valid: being supported by objective truth or generally accepted authority, based on flawless reasoning and solid ground, well grounded, sound, having a conclusion correctly derived from premises, cogent, convincing.

Reliable: suitable or fit to be relied upon, trustworthy, worthy of full confidence, dependable. Campbell and Stanley (1963) have addressed the approaches that can be used to establish the validity of engineering analytical methods and processes. There are two approaches: 1) external, and 2) internal.

External validity is the extent to which the method (approach) is generalizable or transferable. A method's generalizability is the degree the results of its application to a sample population can be attributed to the larger population. A method's transferability is the degree the method's results in one application can be applied in another similar application.

Internal validity is the basic minimum without which the method is uninterpretable. Internal validity of a method addresses the rigor with which a method is conducted - how it is designed, the care taken to conduct measurements, and decisions concerning what was and wasn't measured. There are four different types of internal validity: 1) face, 2) content, 3) criterion-related, and 4) construct.

Face validity is the degree to which a method appears to be appropriate for doing what it intends to do. Face validity is based on justifications provided by the state-of-art and state-of-practice knowledge and experience.

Content validity addresses the degree to which the method addresses the problem (issue) it is intended to address.

Criterion validity addresses the degree to which the method allows for assessment of an issue or problem beyond the testing situation; the generalizability of the method. Criterion validity may be concurrent or predictive; the evaluation may be either be intended to assess a criterion independently evaluated at the same time (concurrent), or to predict achieving a criterion in the future (predictive).

Construct validity addresses the degree to which the results of the method can be accounted for by the explanatory constructs of a sound theory. A method's construct validity can be assessed by specifying the theoretical relationships between the concepts and then examining the empirical relationships between the measures of the concepts, and then interpreting how the observed evidence clarifies the concepts being addressed. Construct validity is demonstrated when measures that are theoretically predicted to be highly interrelated are shown in practice to be highly interrelated.

A reliable method is one that yields valid and consistent results upon repeated use. A reliable method is suitable for its intended purposes. Reliability is established through multiple applications in prototype conditions by independent and qualified users representative of those that will use the method in practice.

Reference: Campbell, D. T. and Stanleny, J. C. (1963): Experimental and Quasi-Experimental Design for Research, Houghton Mifflin, Co., Boston.

back to top ^