Post by ambercyw on Sept 14, 2017 0:21:30 GMT
A bathroom scale that was extremely reliable but not valid would be a scale that measures the same weight every time but it is not the true weight of the object. For example, if I weighed 120 lbs, and the bathroom scale showed that I weighed 150 lbs every time, it is reliable but not valid.
You could not have a scale that was valid but not reliable because in order to be valid, the scale must accurately measure what it is designed to measure. If the measure is valid, it must also be reliable since the weight of the object would be consistent so the displayed measurement on the scale must be consistent as well.
Moghaddam says that a laboratory study could have good reliability and good internal validity but it may still lack external validity. The study could be consistent in measuring the outcome and show the same results at different time points which means it is reliable. The study could also be accurate in determining the independent variable responsible for the outcome (a change in A causes a change in B), making it have good internal validity. However, that same study may not generalize if applied to other individuals; it could be that the laboratory study works for university students in Canada for example but it does not produce the same outcome when it is conducted on university students in China.
The same could be said for an interview or a survey. An interview or a survey could have good reliability meaning that participants would receive the same test score if tested at different times. It may have good internal validity in the sense that it actually measures what it is supposed to measure. But it may lack external validity as it produces accurate test scores for only a certain group of people.
You could not have a scale that was valid but not reliable because in order to be valid, the scale must accurately measure what it is designed to measure. If the measure is valid, it must also be reliable since the weight of the object would be consistent so the displayed measurement on the scale must be consistent as well.
Moghaddam says that a laboratory study could have good reliability and good internal validity but it may still lack external validity. The study could be consistent in measuring the outcome and show the same results at different time points which means it is reliable. The study could also be accurate in determining the independent variable responsible for the outcome (a change in A causes a change in B), making it have good internal validity. However, that same study may not generalize if applied to other individuals; it could be that the laboratory study works for university students in Canada for example but it does not produce the same outcome when it is conducted on university students in China.
The same could be said for an interview or a survey. An interview or a survey could have good reliability meaning that participants would receive the same test score if tested at different times. It may have good internal validity in the sense that it actually measures what it is supposed to measure. But it may lack external validity as it produces accurate test scores for only a certain group of people.