3.1) Deciphering Data
Let’s dive into stats 101! In this section we will discuss the statistics that researchers use to decipher data. The researchers will have already vetted the data, so that you don’t have to. We want you to hear about it, so you can feel confident about the data you are presented with.
Risk is calculated using an odds ratio
An odds ratio (OR) is a measure of association between an exposure and an outcome.
The OR is used to compare the relative odds of the occurrence of the outcome of interest (e.g. condition, disorder, or trait), given exposure to the variable of interest (e.g. having the genetic polymorphism).
It is compared to the occurrence of the outcome of interest (e.g. condition, disorder, or trait), given NO exposure to the variable of interest (e.g. not having the polymorphism).
The odds ratio can be used to determine whether a particular polymorphism is a risk factor for a particular condition, disorder, or trait.
It is also used to compare the magnitude of various risk factors for that condition, disorder, or trait.
Figure 3-1 The odds ratio can be used to determine whether a particular polymorphism is a risk factor and to compare the magnitude of risk factors for a particular disease, disorder, or trait.
OR equal to 1: Exposure does not affect odds of outcome
OR greater than 1: Exposure is associated with higher odds of outcome
OR lesser than 1: Exposure is associated with lower odds of outcome
Q1. When looking at SNPs what is the exposure?
A1. Subject has the polymorphism (e.g. 6% of people).
Q2. What is the absence of exposure?
A2. Subject does not have the polymorphism (e.g. 94% of people).
Figure 3-2 To calculate the odds ratio (OR) divide the odds that a case was exposed by the odds that a control was exposed.
The p-value is used to test the hypothesis
The p-value stands for the “probability value”. It indicates how likely it is that a result occurred by chance alone. The p-value is the basis for hypothesis testing, in which there is a null hypothesis and an alternative hypothesis.
The null hypothesis typically is that there are no differences (no risk), whereas the alternative hypothesis is that there are differences and potential risk.
What does the p-value tell us? The p-value calculates the probability of observing the data (an association) if the null hypothesis (no association) were true. It tells us how often we will be wrong if we infer that there is an association while there is not.
What doesn’t the p-value tell us? It does not calculate the probability whether the null hypothesis is true or not.
What p-value is accepted in general?
In general, a p-value of less than 0.05 is accepted.
If a p-value is larger than 0.05, the data we have observed are not surprising (however we do not know if they indicate an association or not). This doesn’t mean the null-hypothesis is true.