ACCURACY VS. RELIABILITY VS. VALIDITY
“Reliability”, “validity” and “accuracy” – what do they mean?
Reliability and validity are two terms that continue to cause problems for students. Students in Stage 4 and 5 are required to evaluate evidence for reliability and validity. In Stage 6 they are required to discuss and explain how they would improve the validity and reliability of a first-hand investigation. They could also be asked about the validity and/or reliability of secondary sources of information.
With reference to first-hand data, the glossary of terms on pp 76-77 of the Science Years 7 – 10 syllabus defines the terms validity and reliability as follows:
Reliability and validity are two terms that continue to cause problems for students. Students in Stage 4 and 5 are required to evaluate evidence for reliability and validity. In Stage 6 they are required to discuss and explain how they would improve the validity and reliability of a first-hand investigation. They could also be asked about the validity and/or reliability of secondary sources of information.
With reference to first-hand data, the glossary of terms on pp 76-77 of the Science Years 7 – 10 syllabus defines the terms validity and reliability as follows:
When discussing the accuracy of measurements, a dictionary definition such as that provided in the Australian Oxford Dictionary is appropriate:
First-hand investigations
In the context of students planning first-hand investigations, issues relating to accuracy, reliability and validity will impact on the choice of the measuring device and how confident you are about the conclusions drawn from the results of the investigation.
The need for accuracy of data should influence the choice of equipment for conducting first-hand investigations. Where data is collected, quantified or evaluated, reliability refers to the ability of the data gathering process to provide results that are consistent and within expected ranges. It is important that we encourage students to predict expected results and even predict ranges of data.
Validity relates to whether the measurements you are taking are caused by the phenomena you are interested in.
The relationship between reliability and validity can be confusing. Measurements and other observations can be reliable without being valid. A faulty measuring device can consistently provide a wrong value therefore providing reliably incorrect results. However, measurement and observations cannot be valid unless they are reliable and accurate. It is easier to be confident of your conclusion in any investigation when there are limited variables involved and where these variables are controlled.
It is possible to highlight the difficulties involved in such decisions when you think about the difficulty in establishing the link between smoking and lung cancer, and the link between mesothelioma and asbestos dust. The more complex the situation in terms of the range of potential variables that need to be controlled and the difficulty in controlling these variables, the less certain it can be that one test will deliver a valid and reliable answer.
In the context of students planning first-hand investigations, issues relating to accuracy, reliability and validity will impact on the choice of the measuring device and how confident you are about the conclusions drawn from the results of the investigation.
The need for accuracy of data should influence the choice of equipment for conducting first-hand investigations. Where data is collected, quantified or evaluated, reliability refers to the ability of the data gathering process to provide results that are consistent and within expected ranges. It is important that we encourage students to predict expected results and even predict ranges of data.
Validity relates to whether the measurements you are taking are caused by the phenomena you are interested in.
The relationship between reliability and validity can be confusing. Measurements and other observations can be reliable without being valid. A faulty measuring device can consistently provide a wrong value therefore providing reliably incorrect results. However, measurement and observations cannot be valid unless they are reliable and accurate. It is easier to be confident of your conclusion in any investigation when there are limited variables involved and where these variables are controlled.
It is possible to highlight the difficulties involved in such decisions when you think about the difficulty in establishing the link between smoking and lung cancer, and the link between mesothelioma and asbestos dust. The more complex the situation in terms of the range of potential variables that need to be controlled and the difficulty in controlling these variables, the less certain it can be that one test will deliver a valid and reliable answer.
Collecting data from secondary sources
When students have to assess the reliability and validity of information and data from secondary sources, the best procedure is to make comparisons between data and claims of a number of reputable sources.
In determining validity, students might consider the degree to which evidence supports the assertion or claim being evaluated. In some cases, students may be able to make observations or conduct experiments to confirm the reliability of the information they have identified.
Some good questions to ask
When students have to assess the reliability and validity of information and data from secondary sources, the best procedure is to make comparisons between data and claims of a number of reputable sources.
In determining validity, students might consider the degree to which evidence supports the assertion or claim being evaluated. In some cases, students may be able to make observations or conduct experiments to confirm the reliability of the information they have identified.
Some good questions to ask
Copied from: “Curriculum Support for teaching in Science 7 – 12” Volume 11 No. 2 2006 NSW Department of Education and Teaching
The Wikipedia page on Accuracy and Precision may also be useful.
The Wikipedia page on Accuracy and Precision may also be useful.