Statistics Flashcards
What is the difference btw parametric and non parametric tests?
Parametric make assumptions, more powerful and sensitive, and four conditions must be fulfilled to use them : data must be interval/ratio, subjects randomly selected, data normally distributed, variation in results from each condition should be similar.
What does a p-value of >0.05 indicate?
Data is normally distributed. A parametric test can be used.
What does a p-value of less than 0.05 indicate?
Data is not normally distributed and a non parametric test should be used.
Usually means there is a difference btw your groups
Consider clinical vs statistical differences (sample size)
What does a normal distribution curve look like?
A bell shaped curve.
What does a negatively skewed distribution curve look like?
Long tail to negative end of x axis.
What does a positively skewed distribution curve look like?
Long tail to positive end of x axis.
What are inferential statistics used for?
To establish differences or relationships btw variables that can be generalised from the sample to the target population.
What are the two types of inferential test?
Parametric and Non parametric
Name three ways normality can be assessed.
Histograms (not recommended)
Calculate mode, median and mean (distribution curves)
Statistically using tests of normality (p values)
Data are interval or ratio
Normally distributed
Which statistical test would you use?
Parametric as long as not ordinal or nominal data in first place!
Data are interval or ratio
Not normally distributed
Which statistical test would you use?
Non parametric
Describe a Type 1 error in hypothesis testing.
Also know as alpha or false positive - finding a significant difference when one does not exist.
Usually due to choosing the wrong statistical test!
When are type 2 errors made?
Also known as beta or lost opportunity
Commonly done when sample size has been too small
Failure to find a significant difference when there really is one!
Inferential is a word used to indicate that the statistical findings are subject to some presumption in part of the person doing the statistics.
Presented as probabilities- p values
Findings of all quantitative studies are subject to be proved/disproved by subsequent studies.
P values explained:
P= 0.1 same as 10%
The lower the number the better the result.
