ASBHDS III - Session 5 - Lecture 1 - Review and Revision Flashcards
2 - Representative Sample vs. Random Sample
Representative Sample vs. Random Sample
Representative Sample:
We wish to measure something about a large population but to do this would be expensive and impractical so we take a sample of the population whose characteristics are representative of those in the larger population.
Random Sample:
Random samples taken from the population are not intended to be representative of the population. They are based on the principle that each person in a population has an equal chance or probability of being selected for the sample.
3 - Incidence vs. Prevalence
Incidence vs. Prevalence
Incidence:
> Incidence refers to new cases.
> Incidence is a rate i.e. has a time dimension.
> Incidence has units of person-years or per person per year.
Prevalence:
> Prevalence refers to cases both old and new
> Prevalence is a proportion.
> Prevalence had no units.
4 - Ratio vs. Proportion
Ratio vs. Proportion
Ratio:
A ratio is one number divided by another number.
Proportion:
A proportion is a fraction where the top part of the fraction has to be part of the bottom part of the fraction.
5 - Rate vs. Ratio
Rate vs. Ratio
Rate:
A rate is the number of events per population per time.
Ratio:
A ratio is a comparison. The comparison could be rates or other measures such as odds.
6 - Ratio vs. Difference
Ratio vs. Difference
Ratio:
A ratio is a comparison whereby one entity is divided by another entity.
Difference:
A difference is a comparison whereby one entity is subtracted from another entity.
7 - Ratio of Benefit vs. Ratio of Harm
Ratio of Benefit vs. Ratio of Harm
Ratio of Benefit from A w.r.t. B:
A ratio of benefit is the inverse of ratio of harm (i.e. 1/ratio of harm).
Ratio of Harm from B w.r.t. A:
A ratio of harm is the inverse of ratio of benefit (i.e. 1/ratio of benefit).
8 - Crude (unadjusted) vs. Standardised (adjusted)
Crude (unadjusted) vs. Standardised (adjusted)
Crude:
Crude describes what was measured.
Standardised:
Standardised describes what would be measured if the groups being compared were identical in respect of the variables standardised or adjusted for.
9 - Confounding (vs. Bias)
Confounding vs. Bias
Confounding:
Confounding is a feature of the population and becomes an issue when two groups are being compared. It influences the measure of association such that it is liable to misinterpretation if the confounding influences are not taken into account. It should be noted that the unadjusted measure of association is a true description of the association, it is not in any way false. It is the interpretation or inference that is likely to be false.
10 - (Confounding vs.) Bias
Confounding vs. Bias
Bias:
Bias is a feature of the study being conducted. In other words the samples being studied are different or being treated differently. Two broad types are selection bias and information/measurement bias. A biased observation is a false observation in that it does not correctly reflect the population it purports to measure.
11 - 95% Confidence Interval vs. p-value
95% Confidence Interval vs. p-value
95% Confidence Interval:
The 95% confidence interval is the range within which we can be 95% certain that the `true’ value of underlying tendency really lies.
p-value:
The p-value is the probability of getting an observation as extreme as, or more extreme than the one observed assuming that the stated hypothesis is true.
12 - Prospective Study vs. Retrospective Study
Prospective Study vs. Retrospective Study
Prospective:
A prospective study watches for outcomes, such as the development of a disease, during the study period. Cohort studies can be prospective but case-control studies can only be retrospective.
Retrospective:
A retrospective study looks backwards and examines exposures in relation to an outcome that is established at the start of the study. Both cohort and case-control studies can be retrospective.
13 - Random Selection vs. Random Allocation
Random Selection vs. Random Allocation
Random Selection:
Random selection refers to recruitment into a clinical trial and affects external validity i.e. generalisability.
Random Allocation:
Random allocation refers to allocation between the treatment groups in a clinical trial and affects internal validity i.e. whether the comparison is a fair comparison.
14 - Placebo Effect vs. Placebo
Placebo Effect vs. Placebo
> The placebo effect is the effect of believing something of benefit is being done to you.
Since any treatment has both a placebo effect and an active treatment effect, the placebo is used to cancel out the placebo effect of the active treatment in any comparison.
15 - As-Treated vs. Intention-to-Treat
As-Treated vs. Intention-to-Treat
As-Treated:
As-Treated refers to analysis according to what the trial participants did. This assesses the physiological effects of treatments studied.
Intention-to-Treat:
Intention-to-Treat refers to analysis according to the original allocation of the participants. This preserved the original random allocation, reduces bias between treatment groups and reflects what would happen in practise i.e. patients do not always comply with prescribed treatments.
16
16
