2.4. Establishing Measurement Confidence

So, whatever the things that we were talking about yesterday in terms of precision and Accuracy, this come under the issue of whenever you do analysis, chemical analysis, it come under the issue of what is called as quality assurance and quality control or it is called as QA/QC okay. Quality assurance and Quality control, this is a very big part of environmental analysis. The reason this is important is when you ask these questions of precision and accuracy, people are going to make decisions based on the analysis you give.

Quality Assurance (QA) and Quality Control (QC) refer to the processes put in place to ensure the quality and reliability of measurements and analyses in chemical analysis, particularly in environmental studies. When you conduct tests, the results may influence decisions made by policymakers or legal entities. Therefore, having strong QA/QC practices is crucial, as they foster confidence in the data being presented, especially in matters that can lead to legal issues.

Think of QA/QC like a quality check in a bakery. Before selling cakes, the baker checks if each cake is moist, well-decorated, and made with quality ingredients. This ensures that customers receive a high-quality product. Similarly, in environmental analysis, QA/QC ensures that the data provided is accurate and can be used confidently for decision-making.

So, all the amount, the whatever measurement that one gives is indirectly linked to the precision and accuracy and the questions that we asked. So, this quality assurance and quality control in short called as QA/QC is part of an analytical method, okay.

Precision refers to how close multiple measurements are to each other, while accuracy refers to how close a measurement is to the true value. When performing chemical analyses, it’s vital that the results are both precise and accurate because any mistakes can lead to incorrect conclusions and inappropriate decisions.

Imagine you're playing darts. If you always hit the same spot on the dartboard but it's far from the bullseye, you're being precise but not accurate. On the other hand, if your darts are spread out but some hit the bullseye, you are accurate but not precise. In environmental testing, we strive for both accuracy and precision to ensure the data is trustworthy.

So, what we are looking for in the QA/QC is some level of confidence okay. So you have some confidence and this confidence is quantifiable, and this is where you need to use statistics.

In QA/QC, it's important to establish a quantifiable level of confidence in the results obtained through analytical methods. Statistical analysis helps in determining whether measurements meet certain standards, allowing for confidence levels to be defined. For instance, if you report that a pollutant concentration is below a certain limit, you need to statistically assess if this claim holds true given variabilities in the data.

Consider a teacher who grades exams. If she says that 90% of students scored above a 70% on a test, she can show the individual scores and how they compare statistically to prove her point. This level of confidence in measurement helps validate the analysis, just as the teacher's grading demonstrates the students’ performance clearly.

We do the hypothesis testing to do that, and this is defensible. Statistics people have worked on it for a long time. ... So, in statistics also whenever you do this hypothesis testing, you have to make a judgment whether what is acceptable.

Hypothesis testing in statistics involves making predictions or assumptions about a population parameter and then using sample data to determine if there is enough evidence to reject or support that hypothesis. For instance, if there is a regulatory limit for a pollutant, a hypothesis test can be used to determine whether the observed concentration significantly exceeds that limit.

Think of a court case where someone is accused of stealing. The prosecutor must gather evidence to either prove beyond a reasonable doubt that the crime occurred (rejecting the hypothesis that the defendant is innocent) or show that there isn't enough evidence (supporting the hypothesis that the defendant is innocent). This legal process mirrors the scientific method of hypothesis testing.

Whenever you open the instrument, you check, you run one standard, and the standard is falling here and the next day your standard is falling here.

When using analytical instruments, it is common practice to set an acceptable margin of error to account for fluctuations. If a calibration point shifts slightly each day, one must decide an acceptable range for these variations to maintain valid results. Perhaps an acceptable tolerance might be set at 5% or 10%, depending on what is practical and justifiable based on the results needed.

Consider quality control in a factory producing light bulbs. If the factory allows a 10% error margin in light output, it can still ensure that most bulbs work effectively. However, if they relied on stricter tolerances, any small variation might deem many bulbs as defects. In environmental analysis, setting a rational tolerance allows for realistic and usable data.

If I am getting a number, if I get say 100 micrograms per meter cube for some concentration ... this value really maybe between 85 and 115.

Every measurement comes with an uncertainty range, which acknowledges that the reported value might not be exact. For instance, if a concentration is reported as 100 micrograms per cubic meter, this might actually range from 85 to 115 micrograms. Recognizing and understanding this uncertainty helps scientists better communicate the reliability of their findings.

Think of a weather forecast predicting a temperature of 75°F with a margin of error of ±5°F. This means that the actual temperature could be anywhere between 70°F and 80°F. Being aware of this uncertainty helps you dress appropriately for the day, just as knowing the uncertainty in measurement informs environmental decisions.

Environmental QA/QC now becomes a bigger part of your measurement ... you must do a measurement which will stand in court.

The implications of significance in QA/QC are pronounced, especially in the context of environmental data which may be scrutinized in legal settings. Therefore, systematically documenting all QA/QC processes not only strengthens the validity of the measurements but ensures they can withstand legal challenges.

Imagine you're preparing a financial report for a company that predicts revenue. If someone claims fraud, they expect clear documentation confirming every aspect of how the figures were calculated. Similarly, environmental measurements need robust checks to ensure they can prove accurate and defendable under scrutiny.

--