Statistics Flashcards
When do you use a chi-squared test?
qualitative or semi-quanitative data (e.g. percentages of people saying yes vs no)
What tests do you use for normally distributed (parametric) data?
2 groups = students t-test
> 2 groups = anova
Non-normally distributed (non-parametric) data
2 groups = Mann-Whitney U test
> 2 groups = multiple
When do you use a Wilcoxon signed-rank test?
Pre and post intervention, analysing significance of quantitative, non-parametric data?
How do you calculate PPV?
TP / TP + FP
How do you calculate positive likelihood ratio?
sensitivity / (1 - specificity)
How do you calculate the pre-test odds?
pre-test probability (or the prevalence) / (1 - pre-test prob)
if your pre test prob or prevalence is very small (i.e. less than 1%) it will basically be the same as pre-test odds
how do you work out probability from odds?
odds = probability / (1-probability)
How do you calculate the post-test odds?
Post test odds = pre-test odds x likelihood ratio
How do you calculate the post-test probability?
post-test probability = post-test odds / (1 + post test odds)
What are all the steps to working out post test probability?
You need
1) pre test prob (prevalence)
2) Pre-test odds
3) likelihood ratio
4) post-test odds
5) post-test probability
How do you calculate number needed to treat?
if reduce mortality from 20% to 15%?
NNT = 1 / absolute risk reduction
= 1 / (5/100)
= 100/5
= 20
What is type II error?
a wrongful miss
i.e. failing to detect a significant effect when there is one
inherently related to the power of a study which, with a sufficient sample size, can be maximised to reduce the chance of type II error
What is the preclinical phase?
testing of drug in non-human subjects, to gather efficacy, toxicity and pharmacokinetic information
what is phase 0?
pharmacokinetics; particularly oral bioavailability and half life of the drug
What is phase 1?
Testing of drug on health volunteers for safety, involves testing multiple doses
What is phase II?
Testing of drug on patients to assess efficacy and side effects
What is phase III?
Testing of drug on patents to assess efficacy, effectiveness and safety
What is phase IV?
Postmarketing surveillance - watching drug use in public
What is an analytic study?
attempted to quanitfy the relationship between two factors
Effect of intervention or exposure on an outcome
What is a non-analytic or descriptive study?
does not try to quantify the relationship, but tries to give us a picture of what is happening in a population
e.g. the prevalence, incidence or experience of a group
descriptive studies include case reports, case-series, qualitative studies and surveys (cross-sectional) studies, which measure the frequency of several factors, and hence the size of the problem
The main purpose of randomisation is to
Remove bias in the allocation of treatment
What is selection bias?
The 2 groups are intrinsically different due to
- systematic flaws in allocation of patients to each group
Dropouts
Chance