Topic 1 - Design of Experiments Flashcards
LO
LO1 Articulate the importance of statistics in a data-rich world, including current challenges such as ethics, privacy and big data
LO2 Identify the study design behind a dataset and how the study design affects context specific outcomes
Observer bias
Occurs when the participants or investigators are aware of the identity of the 2 groups, so there is bias in responses or evaluations
Selection Bias
- Occurs when some patients are more likely to be chosen over others for the study
e. Hospital selects healthier subjects for surgery.
Placebo
A pretend treatment designed to be neitral and indistinguishable from the treatment
- Patients think they have the treatment
Solution is a randomused double-blind trial
Randomised Controlled Double-Blind Trial (RCT)
- ‘gold standard’ experiment for assessing the effect of a treatment
- Participants are randomly chosen
- patients and investigators don’t know which group they’re in
Types of bias
Selection bias
Observer bias
Consent bias
Survivor bias
Adherer bias
Simpsons paradox
- When there is a clear trend in individual groups that dissapears when the groups are pooled together
- The association between a pair of variables (x,y) reverse sign upon conditioning of a third variable (z), regardless of the value of ‘z’.
Domain knowledge
Domain knowledge is background information that helps us to understand the context and nature of the data.
Bias
Bias is something which effects the ability of the data to accurately measure the treatment effect.
Examples include: selection bias, observer bias and confounding.
Confounding
Confounding (or confusion) occurs when the Treatment and Control Groups differ by some third variable which influences the response that is being studied.
Observational Study
An observational study is one in which the investigator cannot use randomisation for allocation to groups. The assignment of subjects is outside the control of the investigator.
