2. Accuracy in Measurement

So, the question about accuracy that we discussed last class was true value versus measured value. This true value means whatever you are measuring, its actual value that is supposed to be and you are measuring something else. Usually, the idea is to bring the measured value as close as possible to the true value, that's the goal.

In any measurement process, an important concept is the difference between the 'true value' and the 'measured value'. The 'true value' refers to the actual value of the quantity being measured, while the 'measured value' is what an instrument displays when it performs a measurement. The aim is to minimize the difference between the two, thereby improving accuracy. Accuracy is crucial in scientific measurements because the closer the measured value is to the true value, the more reliable your results are.

Imagine you are measuring how tall a building is. If the actual height of the building is 100 meters but your measurement tool shows 98 meters, then the measured value (98m) is different from the true value (100m). Your task is to get as close to 100 meters as possible through careful measurement techniques.

This true value and measured value are not equal, which implies that there are errors associated with the measurement itself and the errors can arise from different sources for various reasons.

Since true and measured values do not match, we have to acknowledge that errors occur during the measurement process. These errors can stem from various factors such as the calibration of the instrument, the technique used in measurement, or external environmental conditions. Understanding these sources of error is vital to improving measurement techniques and enhancing the reliability of results.

Think of a baker weighing ingredients. If the scale is not calibrated correctly, it might show 200 grams when the actual weight is 250 grams. The discrepancy between the true weight and the measured weight highlights the error involved in the process, which can lead to a less-than-perfect cake if not corrected.

So in order to do this, we compare the measured value to what is called a standard. A standard by definition is a very strict thing... So the assumption here is that if the reading that is shown by the instrument is not what the standard is.

To determine the accuracy of a measurement, we usually compare the measured value with a 'standard'. A standard is an established reference point that is known to be accurate. By using this benchmark, we can ascertain whether our measurements are correct. If a measurement differs from the standard, we can conclude that there may be an error in the measurement process or the instrument used.

Taking the earlier example of the weight scale, if you have a 200-gram weight as a standard and place it on the scale, you can verify its reading. If it shows 195 grams, you know there is a problem with the scale because the standard weight confirms that the true value should be 200 grams.

So, this is from very simple measurements like length or mass and all it can go all the way up to very complicated concentration measurements in sophisticated instruments.

Calibration is the process of adjusting an instrument to ensure its measurements match the standards. This is true for all measurements, whether it's measuring length, weight, or even concentration in complex chemical analyses. A well-calibrated instrument provides more accurate and reliable measurements.

Consider a car mechanic tuning a car’s engine. The mechanic uses specific tools (standards) to ensure everything runs smoothly. If the engine runs at a perfect balance as per the manufacturer’s specifications (standard), it leads to better performance and fewer issues down the road.

If I give you a standard 5 gram sample and I say please follow it... So, the simplest thing they can do is they will take it and put it on their balance.

Establishing a standard requires verification and consensus among various users. If someone claims to have a standard weight, others must confirm its accuracy through their instruments. This consensus ensures reliability—if multiple scales show the same measurement for a standard, it reinforces the idea that this standard can be trusted.

Think of it like a group of chefs trying to agree on a recipe. Each chef makes the dish based on their understanding, but they also share results. If many chefs agree that a specific quantity of salt is best for a dish, that measurement becomes the 'standard' for that recipe, ensuring consistency in flavor.

What is called a calibration plot or calibration data... So, the way a standard is usually prepared...

Calibration curves plot the relationship between the quantity being measured and the instrument's response. By taking various standard quantities and plotting their corresponding readings, we can establish a curve that predicts how the instrument will behave with unknown samples. Ideally, this relationship is linear, making it easier to interpret values accurately.

Imagine you are creating a chart that represents hours studied versus test scores. If you plot enough data points and find that as study time increases, test scores also go up in a predictable way, you could draw a straight line. This line (the calibration curve) helps you estimate what score a student might achieve based on how many hours they studied.

So, therefore, when we are doing this, you have to report this, you have to say that it is outside my calibration range.

When reporting measured values based on calibration data, it is crucial to indicate whether these values fall within the validated range of your calibration. If they are outside of this range, the results may not be reliable. This caution helps avoid misinterpretation of data and ensures that decisions based on these measurements are well-founded.

Consider a thermometer used to measure temperature in a laboratory. If the thermometer is only calibrated for temperatures between 0-100 degrees Celsius and you use it to measure something at 200 degrees, you must acknowledge that your reading may be inaccurate and therefore should be treated with caution, as it falls outside of the calibrated range.

If there is a standard established worldwide, you have to prepare your own standards, that will be of course be questioned by other people because it is not accepted worldwide.

Calibration isn't a one-time task; it requires ongoing checks against standards. If no globally accepted standard exists for certain measurements, individuals or labs may develop their own. However, these homegrown standards will often need verification from the broader community to gain credibility. Regular checks ensure that instruments still perform accurately over time.

Think about a fitness tracker that logs your steps. Over time, its accuracy might vary due to wear and tear or software changes. Regularly comparing its step count to a manual count helps ensure it remains trustworthy. Just as you would recalibrate your tracker, labs need to frequently ensure their instruments remain precise.

--