Q-Z of Research: Science to English Dictionary, Part 3

For this definition, let’s look at the simplest possible set up. Imagine you are doing an experiment on whether wearing a tie makes people more confident. You gather a lot of people to take part in this experiment- the people you’ve gathered are your study sample.  You now have to divide this study sample (the people in your experiment) into two groups- an experimental group (that will wear ties) and a control group (that won’t wear ties). If you assigned all the extroverts to one group and the introverts to another, that would bias your results (see bias). If you assigned all the people who prefer formal clothing to one group and all the people who prefer casual clothing to the other, that would also be a source of bias. To ensure that you’re focusing on effects caused specifically by wearing or not wearing a tie, you need to make sure that there is no discernable pattern in whether a person is being assigned to the experimental group vs the control group. If you split your study sample up and randomly assigned individuals to each group, ensuring that there is no set pattern in how people are being assigned, you’d have a randomized experiment. 

(Source: https://isps.yale.edu/node/16697)

If a scientific finding is replicable, that means that different studies looking at the same scientific question all came to similar conclusions. Using our tie analogy from earlier- let’s say you found that wearing a tie makes people more confident. A year later, another group decides to do the same experiment on a different group of people. If they also find that wearing a tie makes people more confident, then your finding is replicable. (Source: https://www.ncbi.nlm.nih.gov/books/NBK547546/)

In an experiment, your sample size is the number of measurements or observations that you made. For example, in our tie analogy, your sample size is the number of people in each experimental group. Say you had 20 people wearing a tie- your experimental sample size is 20. If you had 15 people not wearing a tie, your control sample size is 15. The sample size is an important number for statistical calculations. A bigger sample size (also called a bigger ‘n’) is better and allows us to draw stronger conclusions. 

(Source: https://sciencing.com/meaning-sample-size-5988804.html)

We defined this as part of our definition for ‘double blinded,’ but here it is again!

For this explanation, let’s imagine a checkout counter at a store. You are buying something. I am behind the counter as a cashier. In this case, let’s designate you as the participant and me as the experimenter. Let’s say that this is a very bad store where there are only two types of shirts being sold: t-shirts and tank tops.

In a single-blinded study, you as the participant would not know whether you are buying a t-shirt or a tank top. All you would know is that you’re holding a mysterious brown packet that is a type of shirt, and that you’ve now brought it to the counter to check out. As the experimenter, I would know whether you have a t-shirt or a tank top. While you are ‘blinded’ to which category you’re in, I am not. This is a single-blind study.

In a double-blind study, neither you nor I know what’s in the packet. All I know is that you have either a shirt or a tank top and have come to the counter to check out. All you know is you’ve bought some sort of shirt. In this case, we are both blinded to the category you’ve fallen into- so this is a double-blind study.
In the case of a double-blind study, the actual categories are stored in a database somewhere or blacked out of the data files before scientists analyze them. Single or double-blinded studies help make sure that the results aren’t biased- either by patient behaviour or by the experimental analysis, or both.

(Also see: p value)

When scientist say that their findings are ‘statistically significant’ they mean that the difference observed between the control and experimental group is too big to have happened by chance (or randomly). Scientists don’t just determine this by eye! There’s a whole bunch of statistical tests (for example, the t-test) to determine statistical significance. Going into all of those tests, however, would be an entire 200 level course- so let’s leave it here. 

(Source: https://www.iwh.on.ca/what-researchers-mean-by/statistical-significance)

When a patient is symptomatic, they are showing symptoms of an infection or disease. For example, a symptomatic patient with COVID19 may demonstrate symptoms such as fever, cough and muscle aches.

Unlike a hypothesis, a theory is a statement that is an in-depth and natural explanation of an observed phenomenon.  Theories are usually concise, coherent and applicable in many areas. They can integrate many hypotheses as well. Theories are usually based on careful and rational examination of facts.

Transient means lasting for only a short time. For example, some antibody responses may be transient (antibody levels decline over time), while others are persistent (antibody levels remain over time)

A vector is a carrier for a DNA molecule

The wildtype is the “standard” version of a species as it exists in nature. Their non-wildtype counterparts are mutants, altered forms of the natural species. For example, the wildtype version of the modern goldfish is an unimpressive grey carp. Through human intervention, mutants in the form of delicate gold-orange goldfish are bred.