I come from a school that uses Google Apps for Education, so a lot of the information presented below will be specific to Google Sheets, however, there is a lot of crossover in the functionality of Sheets, Excel and Numbers so hopefully even if you are an Office or Mac person you will be able to gain from the tutorials below.
Processing Data in Google Sheets
Displaying Data on a Lab Report
Precision and accuracy are two words that are easily confused, but their meanings in science have very important differences when used to analyse data.
Precision is the ability to repeat an experiment and get the same results. Accuracy is how close your result is to what could be considered the 'true' result.
The image below is a fairly common way to represent the difference between precision and accuracy if we consider that the middle of the bulls-eye represents the 'true' result.
Identifying Precision and Accuracy in Experimental Results No scientific investigation is ever perfect, so it is the job of the scientist to estimate how imperfect their results are. This is why it is important to be able to understand the precision and accuracy of your investigation. Here are the things to look for:
Precision: 1. The range of your results. Imagine you took five measurements of the temperature of a glass of water. The results were 25°C, 28°C, 22°C, 30°C and 24°C. The lowest temperature measurement obtained was 22°C and the highest was 30°C which gives us a range of 8°C. We are not able to get the same result every time which means our measurements are not precise. 2. The uncertainty of your equipment. Sometimes you might get the same result every time, but it could be because the precision of your equipment is poor. If you use a mass balance that does not measure decimal places, we might get two measurements that both say 50g. But if we were then to measure the same masses on a balance that measures to two decimal places, we could find that one mass was in fact 49.51 and the other was 50.49. This is why we must consider our equipment as an influence on precision.
Accuracy: 1. Closeness to the expected result. Sometimes in a science class students carry out experiments that have been done before in order to learn a concept. This means that an accepted scientific result exists that we can compare our own results to. If, for example, you were conducting an investigation into the chemical formula for magnesium oxide, which we know has a ratio of 1:1, and you obtained a result of 1.3:1, then we can say that your result is not accurate. 2. Correlation to the trendline. If you have a trendline on your graph and the data points do not fit it very well, you can say that your data is not very accurate. However, this does not necessarily mean that data that closely matches a trendline is accurate (see point 3 below for an explanation). It can also be argued that a poor trendline fit might be caused by low precision and not low accuracy. You would need to consider the number of trials in deciding which is the cause (more trials leads to more accuracy). 3. The value of the y intercept. If you are conducting an investigation in which you expect the trendline to pass through (0, 0) and it doesn't this could indicate that your experiment has a systematic error (one that you do the same way every time). This is why a close fit to a trendline does not automatically mean high accuracy.
There are many different ways to identify outliers in your data, and the required method will vary depending on the branch of science you're working in, the type of data you're working with, or even the preference of your teacher. To help simplify the process of identifying outliers in your data, I've created this Google Sheet. By simply copying and pasting your data into this sheet, you will be able to identify outliers in your data using a range of different methods.
Science is based on evidence, and the best evidence is hard data. It is true that data can be manipulated or misrepresented in the hands of the wrong person, but if you have the skills for interpreting data yourself, you will be better positioned to find the truth. When analysing any kind of data, here are some key points to think about:
How accurate is the data?
Is the sample size large enough?
Do the numbers show a significant result?
Have enough variables been controlled?
Does the experiment represent a real life situation?
Do the results of the experiment have applications in multiple areas?
How could the experiment be improved?
To help guide you through these thought processes in more detail, here are four videos that explain the steps to analysing data generated by a student investigation.
Second Hand Data
The list of points above applies when analysing all kinds of data, but when dealing with second-hand data or claims made by people that have conducted their own investigations, there are some other questions that we should ask ourselves:
Who funded the study? Often, studies are carried out or funded by companies that have a vested interest in a particular result. This can cause the people carrying out a study to present their results in a way that is favourable to the company.
Is the claim based on 'clinical research' or peer reviewed science? The term clinical research can be very broadly applied and involve very little reliable science. It is the term that companies often use to mean the tests that they've done in-house. If a study is published in a peer reviewed journal, however, that means the work has been looked over by scientists who are experts in their field. The investigation can then be replicated by other scientists to check that they get the same result.
Is the claim about food? In some countries, such as the United States, foods are not regulated in the same way that medicines are. This makes the process of selling and making claims about a food product is much easier than for a medicine.