probability and distributions

Question 1

what is a normal distribution

Answer 1

• when data is symmetrical around central scores
• i.e., mean, median and mode are equal
• data should fit along a a Gaussian curve

Question 2

how do you calculate skew

Answer 2

3(mean-median)/SD
- if skew is more than 0 it’s negative
- if skew is less than 0 it’s positive

Question 3

what are the tests for distribution

Answer 3

normality tests:
- shapiro-wilk
kolmogorov-smirnov

Question 4

what is a parametric test

Answer 4

tests you run on normally distributed data

Question 5

what does transforming data into Z scores do

Answer 5

• helps us standardise data and reduce the impact of skewness
• tells us exactly how far someone is from the mean

Question 6

how do you calculate a Z score

Answer 6

individual point-group mean/standard deviation

Question 7

what are the pros of z scores

Answer 7

• can transform data to a standardised scale
• scale adheres to normal distribution
• can compare things relative to their own population
• use the entire data set

Question 8

what is a sampling error

Answer 8

• we can estimate values and statistics about a population based on a sample but the value is likely to differ from the true mean of the global population
• we can use standard error to show how sure we are

Question 9

how is the standard error calculated

Answer 9

standard deviation/square root of number of datapoint

Question 10

what does a standard error tell us

Answer 10

• how likely it is our sample will vary from one sampling to another
• the smaller the better

Question 11

what are confidence intervals

Answer 11

• the range of values that, in a certain proportion of samples contain the true value of a statistic
• can be used for visualisation - error bars