Estimation, Confidence Intervals and P-values Flashcards
What does 95% CI mean?
It means you are 95% confident that the value lies between two calculated values.
What does sampling distribution of the mean mean?
It means that different samples taken from the same population will each have a different mean, therefore we look at the distribution of these means
What is the standard error of the mean?
It is the standard deviation of the mean of a group of means (multiple sample means) divided by the square root of the number of people within the sample
What is the calculation for a 95% CI of a sample?
Mean - 1.96 x SE(mean) to mean + 1.96 x SE(mean)
In relation to:
1) Sample size
2) standard deviation
Do small or large values for each of these relate to a closer estimate,action of the mean?
1) Sample size
= larger = estimate closer to the mean
2) standard deviation
= smaller = smaller spread of data = estimate closer to the mean
What is the formula of the SE?
Standards error
SE = SD / √n
N = sample size
How do we calculate the proportion of a population with a disease?
P = r / n
r = number of people with the disease n = population
How do we calculate the SE of a proportion?
SE = √p(1-p)/n
What is the null hypothesis?
No significance / difference exists
Where do p-values come from?
What are they and what values can they take?
They come from significance tests
They are probabilities so must take a value between 0 and 1
What does a p-value ACTUALLY tell you, the probability of what?
The probability of obtaining sample data more extreme than observed if the null hypothesis is true
The likelihood to have observed the difference in a population where no difference exists
If a p-value is >0.05, what can we say about the null hypothesis? Can we accept or reject it?
We can reject it but we can not accept it
What can be said if the CI includes 0?
No difference exists as we must accept that a difference of 0 may exist
In significance testing, what is a type 1 error?
Calculating a significant result
When the null is in fact true
In significance testing, what is a type 2 error?
Calculating a non-significant result
When in fact the null is false
