Introduction to Statistics Flashcards
Statistics - 2
- The science of data
- Collecting, organising, analysing, interpreting and presenting data
Population - 1
- Group of “objects” of which we are looking to gather information on
Census - 1
- Collection of data from every member of population
Sample - 3
- Subcollection of the population
- Different samples may lead to different conlusions about population
- Samples have to be representative and unbiased
Statistical Study Phases - 3
- Prepare
- Analyse
- Conclude
Prepare phase - 3
- Context
- Source of data
- Sampling method
Analyse phase - 3
- Graph data, using appropriate graphs
- Explore data qualitatively and quantitatively
- Apply statistical methods
Sampling Methods - 7
- Voluntary sampler response
- Random sample
- Simple random sample
- Systematic sampling
- Convenience sampling
- Stratified sampling
- Cluster sampling
Voluntary sampler response - 2
- Subjects decide themselves to be included in sample
- Biased
Random sample - 2
- Each member of population has equal probability of being included
- Unbiased
Simple random sample - 3
- Each sample of size n has same probability of being selected
- Unbiased
- Difficult for large populations
Systematic sampling - 2
- After a starting point select every k-th member
- Could be biased, by changing starting point
Convenience sampling - 3
- Choose the most easily available sample
- Biased
- Not a good method but could be useful for first impressions
Stratified sampling - 2
- Divide population in subgroups (strata) such that subjects in same subgroup have same char. then draw a random sample from each group
- Not biased, very representative
Cluster sampling - 2
- Divide population in clusters and randomly select the entire cluster
- Could be biased in small datasets
Variable - 1
- A quantity which can vary
Cause and effect studies terms - 3
- Explanatory variable
- Response variable
- Confounding
Explanatory variable - 2
- Independent
- Might cause the effect being studied
Response variable - 2
- Dependent
- Represents the effect being studied
Confounding - 1
- Influence of different explanatory variables mix and cannot be distinguished
Types of study - 2
- Observational study
- Experiment
Observational study - 2
- Subjects are observed but not modified
- Type of study defined by when the data is obtained
Retrospective Observational - 1
- Data is collected from the past
Cross-sectional Observational - 1
- Data is collected from one point in time
Prospective Observational - 1
- Data has to be collected
Experiment - 2
- A certain treatment is applied to the subjects
- A control and treatment method can be applied
Control and Treatment Experiment - 2
- Single blind, the subject does not know which is treatment and which is placebo
- Double blind, subject and researcher do not know which is treatment and which is placebo
Types of data - 2
- Parameter
- Statistic
Parameter data - 2
- Numerical measure which represents a a characteristic of a population
- Represented with greek symbols
Statistic data - 2
- Numerical measure which represents a characteristic of a sample
- Represented with small letters
Categories of data - 2
- Quantitative (Numerical), numbers
- Qualitative (Categorical), names or labels
Discrete data - 2
- Numerical data
- Number of possible values is countable
Continuos data - 2
- Numerical data
- Collection of values is not countable
Level of measurement - 1
- Nominal
- Ordinal
- Interval
- Ratio
Nominal - 3
- Qualitative data level of measurement
- Cannot be ordered (Names, labels)
- Cannot be used for computations
Ordinal - 3
- Qualitative data level of measurement
- Can be ordered but no meaningful differences
- Cannot be used for computations
Interval - 3
- Quantitative data level of measurement
- No natural zero or starting point
- Ordering and difference between number are meaningful
Ratio - 3
- Quantitative data level of measurement
- There is a natural starting point
- Ordering is possible and differences are meaningful