Design Flashcards
Define: research hypothesis
Statement of researcher’s expectations about the relationship between variables under study.
Define: operational definition
Definition of a variable based on how it will be used in a study; how a DV will be measured; how an IV will be manipulated.
Define: null hypothesis
Statement of no difference or no relationship between variables; statistical hypothesis.
Define: sample
Group of subjects in a study; part of the accessible population; intended to be representative of the whole population.
Define: blinding
Techniques used to reduce experimental bias by keeping the subjects and/or investigators ignorant of group assignments and research hypotheses.
Define: epidemiology
Study of distribution of disease in relation to person, place, and time; measures of risk associated with exposures to disease.
Define: prevalence
Number of existing cases of a disease or condition at a given point in time, expressed as a proportion of the total population at risk.
Define: incidence
The proportion of people who develop a given disease or condition within a specified time period.
Process for addressing a research problem
Sort ideas, facts, theories (in clinical experience and profession literature). Start broad, then narrow down to one specific question.
Primary vs Secondary source in literature
Primary: directly from author, best source of specific information
Secondary: description of 1 or more studies represented by someone other than author (review articles, textbook)
Define: independent variable
Variable that is manipulated or controlled by researcher
Define: dependent variable
Response variable that is assumed to be dependent on or caused by another variable (IV)
Define: sampling bias
Bias that occurs when individuals who are selected for a sample overrepresent or underrepresent the underlying population characteristics
Characteristics of a true experiment (3)
Manipulation of variables
Random assignment
Control group
Design strategies: homogenous subjects
Subjects identical on a potentially confounding variable. (Limitation: results are only generalizable to specific type of people studied)
Design strategies: blocking
Build extraneous variables into study design by using them as IVs. Example: block by age groups
Design strategies: matching
Match subjects based on specific characteristics. Example: age-matched, sex-matched
Design strategies: repeated measures
All levels of the IV experienced by all subjects. Subjects “matched” to themselves.
Threats to validity (4)
Statistical conclusion validity
Internal validity
Construct validity
External validity
Threats to statistical conclusion validity
Low statistical power (small sample size)
Assumption violated
Overall error
Poor reliability of measure
Threats to internal validity
Changes in DV accounted for my other reasons:
History: other events occur between pre/post
Maturation: subjects grew spontaneous healing
Attrition: subject drop out
Testing effects: being tested influences result
Order effects: subject may learn from being tested
Instrumentation
Threats to construct validity
The extent to which the operation represents a construct. (Does the pain scale reflect their pain? Does the stride consistency reflect stability?)
Threats to external validity
Can the results be generalized to people, settings, times other than those in study?
Define: retrospective research
Examination of data collected in the past
Define: prospective research
Measurement collected in the present and the future
Longitudinal research
Follow group over time (+) document change in individuals (-) long term commitments (-) little flexibility (-) threats to internal validity (maturation, history, learning effects)
Cross-sectional research
Study a stratified group at a point in time and draw conclusions (+) less costly and resource-intensive (+) subjects tested only once (-) potential selection bias (-) cohort effects
Vital statistics research
Birth rate: ratio of live births during a specific time period to total population at midpoint of time period
Mortality rate: ratio of deaths in specific period to population at midpoint of time period
