HSC Chemistry Guide:
A complete guide into Accuracy, Reliability and Validity
In this HSC Chemistry Guide, we will explore everything you need to know when answering accuracy, reliability and validity questions in the HSC Chemistry course.
Accuracy
Accuracy relates to how precise is your measurements. The accuracy of your experiment’s results can be affected due to the experimental technique or equipment that you use.
For example, which equipment is more precise, using a medical dropper to estimate the volume of solution added or using a pipette? The answer is pipette as it’s calibrated for precise deliverance of a fixed volume solution.
For another example, if you have a ruler that has reading in increments of 10cm (e.g. 10cm, 20cm, 30cm, etc). How can you improve the accuracy of your length measurements? Well, use a ruler that has more narrow increments! For example, increments in 5cm. This way you can get a more accurate length measurement and therefore a more accurate experiment result.
One more example, which equipment is more precise? A measuring balance that provides a reading to 2 decimal places or one that has 4 decimal places? Hopefully, you said the one with 4 decimal places as it allows you to record a more precise/accurate understanding of the weight.
For experimental technique, you may have chosen the correct equipment but your technique may be wrong! For example, using titration as an example, did you rinse your pipette with the appropriate aliquot solution after rinsing it with water before transferring a new solution into the conical flask?
Another example, for measuring the effect of strong and weak acid on the pH of water, did you calibrate your pH probe then moving to a new beaker of water?
Therefore, by improving the precision of your equipment or experimental techniques, you can increase the accuracy of your results.
The accuracy of an experiment’s result relates how close is your result is to benchmark/published results.
Reliability
When a question is asking you to talk about the reliability of a case study or experiment, you should look at two things.
The first thing is whether or not the experiment or case study has been repeated. Repetition can take in place in many forms. For example, is there a large enough sample size? If the sample size of the experiment is ONE, then you need to look at whether the actual experiment repeated or performed again to obtain another independent set of results. This is because a sample size of ONE basically means there is no repetition.
The second thing about reliability is that you should look at is how consistent are the results or outputs of the case study or experiment.
For example, suppose that you are measuring how much sheeps weigh on average at Taronga Zoo. The five sheeps that you weighed was recorded to be 83kg, 120kg, 40kg, 60kg and 20kg. Would you say your results are consistent? No, they are NOT consistent. This means that if you use the five weights to compute your average weight, your result (average weight of sheep at Taronga Zoo) is not reliable.
The reliability of your experimental results relates to how many unique results/outputs you have and how closely relates they are to each other.
Validity
Lastly, we have validity.
When you are asked to assess the validity of an experiment, you should look at three things.
The first thing is that whether or not the experiment’s method serves to address (relates to) the aim or hypothesis of your experiment. If it does not, it means that your experiment is not valid.
For example, suppose that your experiment is to determine whether or not temperature has an effect on the solubility of a compound. However, in the experiment’s method you are measuring the weight of sheeps. Clearly, the method is nothing to do to address the aim of your experiment. Therefore, the results of your experiment WILL NOT be valid.
Secondly, you will need to see if the independent variable and dependent variable are correct. This is sort of similar to the first condition of seeing if your method actually help address the aim of the hypothesis.
Thirdly, you need to ensure that controlled variables are in place so that the independent variable is the ONLY variable that affect the dependent variable.
Fourthly, depending on the experiment, you may need to verify that you have accounted for all variables (under the controlled variables group) that may affect the dependent variable by setting up a control.
A control is basically like a separate, mirror (identical) experiment except you do not ADD your independent variable.
For example, if you are measuring how much carbon dioxide has escaped due to heating the coke for 15 minutes. This means you need to remove the amount of carbon dioxide that escapes the coke naturally (without heating). To do this, you should set up an appropriate control for this experiment.
A control for this experiment could be having a open coke can exposed to the same environmental conditions except you do not heat the coke!
The difference in the mass of the coke will be due to carbon dioxide escaped due to heating.