Statistics Flashcards Preview

Paper 3 > Statistics > Flashcards

Flashcards in Statistics Deck (18)
Loading flashcards...
1
Q

t test

A

For comparison of two means.
Assume the data is parametric.
Sample size >=3.

2
Q

Paired t test

A

Comparison of two means where the data is derived from repeated measures.
Assumes parametric data.
N>=3

3
Q

Mann-Whitney

A

The non-parametric equivalent of the t test.

4
Q

Linear regression

A

f

5
Q

SEM

A

Standard error of the mean.
Rather than being descriptive like S.D. , SEM gives an appreciation of how close your sample mean is to the population mean.

6
Q

SD

A

Describes the variance/spread of your sample.

7
Q

95% CI

A

There is a 95% chance that the population mean lies within the range.

8
Q

P value

A

Used to decide whether a particular data point was significantly different from the control value. Under the null-hypothesis there is less than 5% probability that the control and treated data are random samples from the same distribution.

9
Q

ANOVA one way

A

Analysis of variance.
Used to compare multiple means with only one independent variable.
Assumes the data are parametric and n>3.

10
Q

ANOVA two way

A

Comparison of multiple means where there are two independent variables.
Assumes same as ANOVA one way.

11
Q

Tukey’s post hoc

A

A

12
Q

Dunnet’s post hoc

A

f

13
Q

Kruskal-Wallis test

A

f

14
Q

Wilcoxon test

A

f

15
Q

Friedman’s test

A

f

16
Q

Pearsons

A

This is for linear “least squares” regression. Pearon’s ”r” value is the correlation co-efficient derived by the product moment method. A value close to 0 represents a lack of correlation, whereas a value of +1 or -1 represents a perfect positive or negative linear correlation. The ”r2” value is the square of this “r” value and represents the “co-efficient of determination”. It ranges between 0 and 1 with 1 representing a perfect linear correlation between two variables.

17
Q

Spearman

A

f

18
Q

Logrank test

A

The logrank test is a non-parametric test for statistical significance testing of data that is not normally distributed.