What is the nicest example that shows the difference between validity and reliability?
One of the most commonly cited examples to illustrate the difference between validity and reliability is that of a target and an archer.
Imagine a target that is divided into several concentric circles, with the bullseye being the center. A reliable archer would consistently hit the same spot on the target with each shot, even if that spot is not the bullseye. On the other hand, a valid archer would hit the bullseye or the center of the target, which is the intended target.
In this example, reliability refers to consistency or repeatability of results, while validity refers to accuracy or whether the results measure what they are supposed to measure. The archer who consistently hits the same spot on the target demonstrates reliability, while the archer who hits the intended target or the bullseye demonstrates validity.
Another example could be a bathroom scale. A reliable scale would give you the same reading each time you step on it, even if the reading is not an accurate reflection of your actual weight. A valid scale, on the other hand, would give you an accurate reading of your weight.
It is important to note that both reliability and validity are essential for any measurement instrument or assessment tool. A reliable instrument that consistently produces inaccurate or invalid results is of little use, and a valid instrument that produces inconsistent or unreliable results is also not very useful.
Paul Hartog In social science research methods reliability is defined as the ability to replicate a piece of research and come up with the same results. It can also be described as consistency in findings.
Validity is about measuring what you set out to measure, and sometimes labelled 'truthfulness'.
Each has value. Reliability is essential for quantitative studies, while validity is essential for qualitative studies. it is rare for any piece of research to be both reliable and valid as different methods are required for each. However, use of multiple methods goes some way to achieving this.
Without getting into jargon or intent in specific disciplines, just take the definitions. "Valid" means correct or true. It may occur commonly or not. Yet, nothing can change the "validity" (true or false) of something. "Reliability" means getting the same results for the same inputs, whether "valid" or not. You can reliably get invalid information (e.g. most current news). Reliability can cover a continuous range from totally unreliable (random) to totally reliable (100% predictable). Validity is binary. Something is either valid, or not. If there is perceived ambiguity, you have not "granulated" your question or subject enough.
Steven Cooke I am not inclined to completely agree with your analysis. I don't agree that validity is binary. This is a somewhat harsh interpretation of its meaning.
Valerie Saunders , yes, I made it a bit strict, but both terms may have too much "range" otherwise. Strictly speaking, "valid" derives from Latin for "strong", so there certainly could be a relative scale of the amount of strength in an argument, less so with any facts supporting it.
I have always recommended to students and colleagues that important terms in any communication - and particularly research - be defined precisely at the start. Many words have different uses in different environments, and even in the same discipline by some people.
Clarifying what YOU mean by any term used is the best way to avoid ambiguity.
The real problem is when we start to mix validity with reliability on continuous scales (sometimes we have to, in risk analysis). NEITHER should be a synonym for "Fact".
Steven Cooke HI Steve, thanks for the message. Please note that I never mix validity with reliability, and teach both concepts as being completely separate.
- Badges
- Science topic