Definition of biostatistics
The science of collecting, organizing, analyzing, interpreting and presenting data for the purpose of more effective decisions in clinical context.
4 Importance of biostatistics
IDID
- Identify and develop treatments for disease and estimate their effects
- Design, monitor, analyze, interpret, and report results of clinical studies
- Identify risk factors for diseases
- Develop statistical methodologies to address questions arising from medical/public health data
When do you need biostatistics?
BEFORE you start your study!
dx betw population and a sample
- Population includes all objects of interest
- assoc w/ PARAMETERS(μ,σ)
- Sample is only a portion of the population
- assoc w/ STATISTICS(X,s)
compute statistics, and use them to estimate
parameters.
reasons why we dont work with populations
- usually large, and often impossible to get
data for every object of study - Sampling is costly, and the
more items surveyed, the larger the cost
Descriptive vs Inferential statistics
statistics are computed in order to estimate the parameters of a population
Descriptive Statistics
- first (computational) part of statistical analysis
- procedure used to organize and summarize masses
of data
Inferential Statistics i
- second (estimated) part of statisticcal analysis
- used to find out info about a population, based on a sample
define biased sample
Biased sample is one in which the method used to
create the sample results in samples that are
systematically different from the population.
define random sampling
Each element/item in the population has an equal chance of
occuring.
- preferred way but difficult to execute
- requires complete list of each element in pop therefore usually assoc w/ comp gen list
define systematic sampling
elements are counted off /every x-th element is taken
OTHER TYPES OF SAMPLING
convenience sampling:
readily available data is used (first people the surveyor runs into.)
cluster sampling:
- divides the pop into groups/clusters usually geographically.
- clusters are randomly selected,
- each element in the chosen clusters are used.
Stratified sampling
- divides the population into groups called strata.
- by some characteristic,(M/F) not geographically
- sample taken from each strata using
- random, systematic, or convenience sampling.
3 things thtat determine a good sample
- Random selection
- Representativeness by structure
- Representativeness by number of cases
types of error
Random error = sampling variability.
Bias (systematic error) difference betw/ observed value and the true value due to all causes other than sampling variability.
absence of error of all kinds = accuracy
sample size calculation principles
- law of large numbers= as the sample size increases the margin of error decreases as percentage diff betw/ popo and sample goes to zero
- number of experimental units are justified
- purpose of size calculation = large enough for acc info but small enough for practicality
factors sample size depend on APEUS
Acceptable level of confidence
Power of the study
Expected effect size
Underlying event rate in the population
Standard deviation in the population
stages of biomedical research
- Planning and organization
- Conduction of the investigation
- Data processing and analyses of results
8 components of research programme
- Aim : summary and
formulatation of the research hypothesis. - Object: event, that is going to be studied.
- Units of observation: logical(studied case) and technical(evn of logical unit
- Indices of observation: important mesaurable factors. they are *measurable*additive*self controlling
- factorial
- resultative
- Place
- Time:
- single: studied in single “critical’ moments
- continuous: show long term tendency of events
- Statistical analyses
- Methodology
one vs many measurements
many measurements on one subject: get to know the one subject quite well but learn nothing about how the response varies across subjects.
one measurement on many subjects, you learn less about each individual, but you get a good sense of how the response varies across subjects.
explain paired and unpaired data
paired: 2+ measurements are made on the same observational unit (subjects, couples)
unpaired: only one type of measurement is made on each unit.
describe the parts of the research plan in planning and organisation
- Definition of the team responsible for the study and preliminary training.
- Administration and management of the study.
key components of information processing
- Data check and correction
- Data coding
- Data aggregation
- according to data use: Primary /Secondary
- according to indice number: Simple /Complex
benefits of data summary
- become familiar with the data and the characteristics of the sample that you are studying
- identify problems with data collection or
errors in the data (data management issues) - Range checks for illogical values
dx betw/ variable and data
variable: something whose value can vary.
data: values obtained from measuring a variable
categories of variables
Nominal
- Values in arbitrary categories
- Order of the categories is completely arbitrary/ meaningless
- No units!
- Data has no units of measurement.
Ordinal
- Values in ordered categories
- Order of the categories is not arbitrary. possible to
order the categories in a meaningful way. - No units!
- Data do not have any units of measurement.
what are the four levels of measurement
- Nominal is the lowest level. Only names are meaningful here.
- genotype- no numbers calc is meaningless
- Ordinal adds an order to the names.
- pain score-order matters but not diff betw
- Interval adds meaningful differences.
- temp- diff betw 2 points
- Ratio adds a zero so that ratios are meaningful.
- height- clear def of 0, can look at the ratio of 2 measurements
