CE Flashcards
Name the types of variables
Numerical
Categorical
Can a variable be both a numerical and a categorical?
Yes numerical values can be put into categories, however categorical values cannot be changed to numerical value
Name the variable 2 types of numerical variables
Continuous variable
Discrete variable
Which variable is presented in a frequency table
Categorical variable
Name the 2 types or Categorical vaiables
Ordered categorical
Unordered categorical
What is a binary/ dichotomous variable
This is a type of unordered categorical variable that only has 2 groups.
When do we use a histogram rather than a bar chart?
When the variable is numerical
Bar charts and frequency tables are used for categorical data
What is another term for RISK?
Cumulative incidence
What is the difference between risk and prevalence
Prevelance shows the amount of disease at a particular time whereas the risk talks about the number of new cases over a period of time
What is the equation for 95% reference range?
mean +/- ( 1.96 X Standard Deviation)
What is an interventional study
This is where the investigator assesses whether changing/ modifying something affects the progression of a patients disease.
Name 3 observational studies
Case-control Cohort study ecological descriptive cross sectional
Define accuracy
Relates to how representativethe sample is to the true population
Define Precision
Relates to the amount of variation between the sample
Name the types of bias (internal validity)
Selection bias
Measurement bias
Define Standard deviation
A measure of the
spread of observations around the mean
Define standard error
Measures the precision of the sample mean as an estimate of the populaton mean
What is the equation for SEM?
SEM= SD / √N
What is the equation for 95% Confidence interval?
CI = (statistic) +/- (1.96 x SE)
When do you use a Reference range and not a Confidence interval
RR measures the variation between individual observations in the sample, whereas CI shows the variation in a group of individuals
List the 3 criteria for a variable to be a confounder?
- Associated with the exposure
- Be a risk factor for the disease
- Must not be on the causal pathway between disease and exposure
On what occasion would the Risk ratio and Odds ratio be similar
When the disease is rare
Define case-control study
When individuals are recruited on the basis of their outcome/ disease and then their past exposure is investigated AND a control group is picked with similar properties but no disease
Name the types of cohort studies
Historical
Prospective
What is a prospective cohort study?
When a sample of people with a certain criteria are picked and then followed up over a couple of years to see if the outcome develops.
What is a historical cohort study?
Subjects who already have the outcome are selected and researchers try and investigate their past exposure. (No control group with no disease)
How can a variable be classed as a confounder?
- Associated with the exposure
- Risk factor for the disease
- Must not be on the causal pathway between Exposure and disease
How to remove confounders?
- Randomisation-
2. Exclusion: use subjects from the same criteria- e.g same social class
What is a Multivariable model?
It’s a model that takes into account the effects of many different confounders
Name the models that combine association estimates after stratification.
- Mantel-haenszel methods
2. Regression models
What are correlation and linear regression used for?
Models that are used to investigate the association between two numerical variables
What is the regression line?
The line/equation which best describes the relationship between 2 variables.
