Week 6- Introduction to Evidence Flashcards
What is a correlation?
Technique for investigating the relationship between two numerical values
What is a correlation coefficient?
Measurement of the relationship between two numerical variables/measurements
What is a positive correlation?
High values in one variable are associated with high values of another variable
What is a negative correlation?
High values on one variable are associated with low values of another variable
How do you measure the strength of the correlation?
Close to 1 or -1: strong relationship
Describe the Pearson correlation coefficient
- Measures strength of linear association between 2 quantitative variables
- Below 0= negative correlation
- Above 0= positive correlation
- Sample size above 30
Describe the Spearman correlation coefficient
- Measures strength of association between two ranked variables
- Sample size small (below 30)
What is the difference between parametric and non parametric data?
Parametric: assumes normal distribution, draws more conclusions from sample to population
Non-Parametric: no assumption on distribution, simple and less effected by outliers
What is a hierarchy of evidence?
- Ranking system, ranks quality of research
- What study design provides most robust evidence
- Based in rigour and ability to minimise bias
- Understanding ability of studies to minimise bias
What is the purpose of hierarchy of evidence?
- Helps to be a discerning consumer of research (can I trust research?)
- Some research designs better at reducing chance, minimising bias, control confounders
- Choose best evidence to answer question
What are examples of reporting bias?
- Publication Bias
- Location Bias
- Language Bias
What is publication bias?
Selection of research evidence due to positive findings published in some areas
What is location bias?
Only selecting research evidence from one source
What is language bias?
Selecting research evidence from you native language
What are examples of methodological bias?
- Sampling Bias
- Allocation Bias
- Maturation Bias
- Attrition Bias
- Measurement Bias
- Placebo
- Hawthorne Effect