However when looking at reliability in real life situations this is not always the case. It would therefore appear that if there is no error and the results are consistent it must also be valid. Less random error in the variables means results won’t be affected by differences in the variables so results should be consistently similar. The more reliable results are the less random error there is. This diagram shows that reliability can have problems with precision whereas validity has problems with accuracy. The red circle in the middle represents the ideal measurements you want to be taking and as you get further from the circle the results get less relevant to your research question as they are not measuring it. This diagram is an easy way to visualise how validity and reliability are different. Validity scientifically answers the question it is intended to answer. Reliability is measuring a dependent variable and getting consistent results. “Reliability is necessary but not sufficient for validity.” What I believe this demonstrates is that reliability is necessary if you wish to conclusively prove a hypothesis but it is null or void, if the test lacks validity because otherwise you are unable to infer anything from your results about the question you asked.
