Biostatistics Flashcards
Statistics
- Study of how information should be reflected on
- Given guidance for action in a particular situation involving uncertainty
Biostatistics
- Application of statistical methods to the medical and health sciences
- Includes epidemiology
Main Goals
- To obtain descriptive information about the population from which the sample is drawn
- To test research hypothesis about population
Population
Any large collection of objects or individuals about which information is desired
Parameter
Summary number that describes the entire POPULATION, averages or percentages for example
Sample
Representative group drawn from the population
Statistic
Any summary number that describes the SAMPLE, like an average or percentage
Population v.s. Sample
- Population: Contains all members of a specified group
- Sample: part/subset of population, ALWAYS less than entire population
Why sample?
- Economic advantage
- Time factor
- Very large populations
- Inaccessible populations
- Destructive nature of the observation
Types of Data
- Binary (discrete)
- Categorical (discrete)
- Continuous
Binary Data Examples
- Yes/No
- Success/failure
- Alive/dead
Categorical Data
- Nominal: unordered/qualitative
- Gender, race, martial status, education, etc.
- Ordinal/Hierarchial: scales or statuses for comparison
Continuous
- Age, height, weight, temperature, distance, etc.
- Measured on continuum or scale
- Ratio accomadates a value of zero
- Distance between each unit has the same meaning or measurement
- Continuous data can have almost ANY numerical value and can be meaningful at any interval
Mean
- Arithmetic average
- Measure of central tendency for continuous variables
- Affected by outliers
Medium
- Middle value of an ordered data distribution (50th percentile)
- Measure of central tendency for continuous variables
- NOT affected by outliers
Mode
- Most frequent value
- Not affected by outliers
- Measure of central tendency
Variability
- Describes data dispersion
- Helps define whether study groups are drawn from different populations
Range
- Measure of variability
- Difference between max and min values
Quartiles
- Measure of variability
- Q1 and Q3 values separate the bottom and top 25% of the data
Standard Deviation
- SD
- Measure to quantify dispersion/variation of a set of values
- Quantified the difference of individual observation from teh mean of the set values
Standard Error of the Mean
- SEM
- Standard deviation of the sampling distribution of the mean
SEM v.s. SD
- SEM describes how precise the mean of the sample is compared to the true mean of the population
- Sample size increases, SEM decreases
- SD may be more or less as sample size increases depending on the dispersion of the additional data added to the sample
