Bio fucking stats :( Flashcards
Why do we obtain a sample?
We obtain one representative sample and use this to draw conclusions/ inference about the population.
Why don’t we sample the whole population?
Can’t sample whole population as it is to costly and difficult to investigate.
What are categorical variables?
Binary variable e.g. colour of m&ms, type of bone fracture etc..
Summarise by the proportion/ percent
What are continuous values?
Can take on any value e.g. height, weight, age, blood pressure
Often interested in mean or average value
What is a mean?
Sum of all variables divided by the number of observation
How do you study a continuous variable?
Have a population and take a sample - record this mean.
Record in sampling distribution (e.g. take sample of 30, record the mean and record this on the graph)
Taking a sample mean of 30 vs 100 (100 has more certaincy, so distribution is narrower)
What is the standard deviation?
Looks at every point and works out how far it is away from the mean (does this for every point)
- Works out average spread/ distance of those points away from the mean.
- Measure ‘average deviations’ away from the mean
What are the two different sorts of errors?
- Errors that make our answers more uncertain i.e. more variability (meaurement error)
- Errors that move us away from the truth i.e. we get the wrong answers often called bias
Can’t avoid 1. Really important to avoid 2 (we don’t want a biased answer)
How can we avoid Error 2?
A random sample from the whole population (as long as everyone takes part) can avoid 2.
What is out sample supposed to match?
The target population but sometimes may get people involved which are not from the target population.
Why does bias matter?
In medicine studies need to check all possible angles to see if the drug is safe to consume.
