Last min stats cards

Question 1

for asymmetrical tests what short of stats should be used

Answer 1

stats which are insensitive to extreme values

Question 2

SE

Answer 2

tells us how far on average the sample estimates of the mean would be from the true mean, if you carried out this study a large number of times , using different samples of the same size from the same population

Question 3

Odds=

Answer 3

number of positive cases/ number of negative cases

Question 4

Odds ratio

Answer 4

Odds in group 1/ Odds in group 2

Question 5

define Odds

Answer 5

the ratio of likeliness of an event happening compared to the likelihood to it not happening
- can range from 0- infinity

Question 6

define risk

Answer 6

the probability of having a disease in a group

Question 7

risk=

Answer 7

no. of people with disease/ total number of population

Question 8

Odds- ratios

Answer 8

odds/ odds

e.g. odds that disease group exposed/ odds that control group exposed

Question 9

risk ratio

Answer 9

probability/ probability

e.g. probability of outcome if on drug/ probability of outcome not on drug

Question 10

when can odds ratios be used

Answer 10

case-control

Question 11

when can risk ratio be used

Answer 11

cohort/ act (agRRrrrr- group of pirates)

Question 12

RR or OR = 1

Answer 12

no difference

Question 13

risk difference

Answer 13

absolutely difference
0= no difference
- difference between score in group

Question 14

non parametric stat compare

Answer 14

entire distributions instead of means

- insensitive to extreme values

Question 15

Mann- whitney

Answer 15

non-parametric test used to quantify differences in distributions between two group

Question 16

wilcoxon

Answer 16

non-parametric test used to quantify differences in distributions between two paired groups

Question 17

Kruskall wallis

Answer 17

non-parametric test used to quantify differences in distributions between more than three independent groups

Question 18

Friedmen

Answer 18

non-parametric test used to quantify differences in distributions between more than three paired groups

Question 19

positives of non-parametric

Answer 19

• can be used when data is skewed/ asymmetric
• can be used for smaller sample sizes
• provides similar p values to parametric tests

Question 20

negatives of non-parametric

Answer 20

• doesn’t compare means, but total distributions
• no means, p-values or CIs provided
• less sensitive to extreme variables

Question 21

correlation

Answer 21

the extent to which one variable is associated with another variable
- how much one variable relies on the other

Question 22

pearsons

Answer 22

parametric

Question 23

spearmint

Answer 23

non-parametric

Question 24

persons is a

Answer 24

correlation coeffieicinet which quantifies the strength of association between two variables which have a linear relationship

Question 25

spearmint is a

Answer 25

correlation coefficient which quantifies the strength of association between two variables which have a non-linear relationship

Question 26

R2 is the

Answer 26

coefficient of determination and quantifies the proportion o f variation in one variable which can be used to explained another rvariable

Question 27

regression is a

Answer 27

mathematical equation used to describe the linear relationship between quantitative outcome and quantitive predictors

Question 28

regression equation

Answer 28

outcome= a x b x predictor

Question 29

a

Answer 29

constant= intercept

- mean value of outcome when predictor is zero

Question 30

b

Answer 30

slope

- predicted increase int he outcome for each one unit increase in the predictor

Question 31

regression assumptions

Answer 31

linear regression
residuals are normally distributed
homoscedasticity ( constant variance)

Question 32

homoscedasticity

Answer 32

variability in the residuals is the same across the predicted values

Question 33

chi squared

Answer 33

parametric

• needs at least 40 participants
• or if between 20-39, if 5 values in each cell

Question 34

if assumptions don’t apply to chi squared

Answer 34

fishers exact test

Question 35

assumptions of fishers

Answer 35

• non parametric
• fewer than 20
• less than 5 in each box if between 20-39