We explain what reliability is and how it is related to quality and validity. In addition, we tell you what a reliability study is.
What is reliability?
Reliability or trustworthiness is, in certain fields of knowledge, the greater or lesser tendency to obtain errors in a given process, especially with regard to measurement and research through instruments or procedures. In other words, the less prone an instrument, procedure or methodology is to error, the greater its degree of reliability or trustworthiness.
The term “reliability” comes from Latin fidestranslatable as “faith”, “trust” or “loyalty”, and the additions with- (“next to”) and -bile (“that can”). So reliability can be understood as the possibility of trusting something, understood as a property of the object and not of the trustor. For example, a computer brand has high reliability if its machines turn out to be of good quality, efficient in their use and, therefore, worthy of the buyer's trust.
Unlike other properties, reliability is normally determined by an evaluation period : a period of time in which the effectiveness or accuracy of an instrument, whether a machine, system or procedure, is evaluated and during which it must perform the required functions. for this Proof tests are used, intended to produce a reliability coefficient that is, a score that reflects how trustworthy an instrument is.
Reliability is an extremely important concept in the industrial world, whose processes require high automation and a low margin of error. In fact, reliability is sought not only in instruments and production circuits, but in companies as a whole and in their associated brands.
Quality and reliability
In the contemporary world, reliability and quality are concepts that are often closely related: The quality of objects and instruments is normally measured and expressed through their reliability. . This means that, in general, a high-quality object will tend to be more reliable, since better materials or a more rigorous and demanding testing process have been used in its manufacture. Similarly, low quality objects are shown to be unreliable, that is, they may fail in a significant percentage of cases.
Reliability and validity
In the field of research, measurements can be more or less exact, and therefore more or less reliable, depending on how many times they yield accurate results. So, In the study and testing of measurement methods, two important variables are normally identified:
- Reliability: how trustworthy the measurement is.
- Validity: how useful it is to get to the truth.
There is an important relationship between both concepts: one determines the other. A measurement will be reliable whether it is reasonable to assume that their results are accurate, since most of the time they tend to be. Similarly, a measurement will be valid if it yields true information about the measured reality.
For example, an exit polling methodology will be more or less reliable, depending on whether its results are more or less similar to those produced by the final election. That is, if your perceptions turn out to be close to the truth, it will be more reliable and, therefore, more valid, since with its results you can think about the electoral panorama with a low margin of error. On the other hand, a survey on consumer habits would not be valid to think about the electoral issue, since it measures other issues of reality that are not of interest in this context, even if they are very reliable.
What is a reliability study?
A reliability study is called evaluation or analysis of the risks of failure of an instrument, a system or a team of individuals . These types of studies are usually carried out in very different contexts, always taking into account the technical or specialized elements that determine failure or success, in order to establish some type of statistical margin of failure: a statistical prediction of the tendency for equipment to fail. or instruments.
Reliability studies They allow you to identify critical elements or frequently failed components, understand failure dynamics and thus improve reliability. of processes, equipment or instruments. In fact, they are part of product quality and risk control studies, as well as maintenance or preventive maintenance plans.
Reliability of an investigation
The reliability of a research is, whatever its context or objective, the degree of confidence that can be placed in its results . To determine the reliability of research (something crucial in the development of science), various methodologies and basic evaluation principles are usually applied to determine three fundamental aspects of its reliability:
- The internal consistency that is, that its fundamental elements do not contradict each other. This can be determined, in turn, using methods such as piecewise reliability tests, Kuder-Richardson coefficient, or Cronbach's alpha coefficient.
- stability that is, obtaining similar results in repeated tests in parallel (test-retest tests). The more times similar results are obtained, the more stable the research will be and, therefore, the more reliable.
- The equivalence that is, observation and analysis between peers, applying specialized criteria to validate the result obtained and, therefore, also validate the method with which it was obtained.
References
- “Reliability (psychometrics)” on Wikipedia.
- “Reliability” in the Language Dictionary of the Royal Spanish Academy.
- “Radication of the word Reliability” in the Online Spanish Etymological Dictionary.
- “Reliability and validity of research instruments” by Roberto Marroquín at the Enrique Guzmán y Valle National University of Education (Peru).
