Statistics Flashcards
What is another name for normal distribution
Gaussian distribution curve
What is another name for normal distribution
Gaussian distribution curve
Why is normal distribution important
Followed by many biological variables
E.g bottom of curve it’s uncommon feature
Top of curve most common or average
End of curve less frequent but still common
The mean is
Measure of location (average)
Standard deviation is
Measure of dispersion (variability)
Mean and standard deviation can be used to calculate what
Equation of the curve
Why is calculating the equation of the curve useful
To calculate the population
Subtracting one standard deviation from the mean and adding one standard deviation will be
68 percent of people will be within one standard deviation of the mean.
Multiply standard deviation by 1.96 and add and subtract that from the mean.
Predictions for normal distribution
50 percent of values lie below and above the
95 percent of values lie between the mean Substrat 1.96x standard deviation
This means 95 percent of values between mean subtract 1.96xsd and mean add 1.96xsd.
You only use the mean and standard deviation if the data is…
Normally distributed.
Population
Entire group that we would ideally like to study
Sample
Sub group of population we actually study
Example
Population of interest;all men over 18 living in the uk
Sample: volunteer males aged over 20 working at factories in five cities in the uk
Is this representative
No
Population value
Is the population mean
Example mean height of all men in the uk
Sample value
Is sample mean
If sample is representative this is good unbiased estimate of the population mean
Two main factors affect the size of this uncertainty
Sample size the more people we study the more confidence we have in estimates accuracy
Standard deviation (variability)
Two related measures of uncertainty
Standard error- quantifies uncertainty of sample ,am as an estimate of the population mean
Confidence interval provides margin of error of sampl mean as estimate of the population mean.
Standard error of mean
Smaller than sd of individual values
What is confidence level
How sure we are that are results are accurate or not
Standard error
Tells us how precise accurate our sample mean is as an estimate of the population mean.
Calculate by dividing the standard deviation by the square root of sample size
Used to calculate 95 percent confidence intervals expect 95 percent of sample means to fall within -+ 1.96SE of population mean.
What can you calculate a confidence interval for
What statistics
Mean median risk proportion odds ratio
How to calculate incidence risk
Developed disease/ total
Total= developed disease and did not develop it
Null hypothesis definition
There is no real difference between groups
Null hypothesis definition
There is no real difference between groups
Conduct study and collect data
Perform a statistical test and calculate a test statistic
Convert test statistic into a p value
Interpret p value and draw conclusions
Conduct study and collect data
Perform a statistical test and calculate a test statistic
Convert test statistic into a p value
Interpret p value and draw conclusions
If p<0.05
Reject null hypothesis less than 5 percent probability of obtaining something
What is the starting point of statistical analysis
Null hypothesis
Need to mention word ‘true’ or ‘in the population’
Example: the true mean FEV1 for men is the same as the true mean FEV1 for women
Alternative hypothesis:the true mean fev1 for men is different to the true mean fev1 for women.
Take sample 39 men and 46 women
Observed difference in mean fev1 is 1.34 litres
Males mean 4.651 SD=0.761
Females mean=3.31 SD=0.657
P less than 0.05 small
P>0.05 large
If p is larger than 0.05
Not enough evidence to reject the null hypothesis at 5 percent significance
T value
Mean difference/se of mean difference
