Physics and statistics Flashcards
What is the independent variable or dependent variable?
Scrivani. Invited review - study design considerations. VRU 54.4 (2013).
Independent - imposed by researcher - attribute that is believed to predict or cause fluctuation in a dependent variable. AKA: event
Dependent - attribute that is not under control fo researcher - outcome of or response to independent variable. AKA: response to the independent variable.
What is the difference between different studies: experimental vs observational studies.
Scrivani. Invited review - study design considerations. VRU 54.4 (2013).
Observational: collect information about attributes or measures of group of subjects, but do nothing to affect subjects: epidemiolgoic studies. Example: US findings in cats with mast cell disease. Cannto claim that these findings are specific to mast cell disease, because they lacked a control group.
Experimental studies: investigate effects of an itnervention by the researcher to the individuals in a sample. Example: describe new techniques r treatments that have not been compared to a control group of subjects.
What is the difference between longitudinal or cross-sectional studies?
Scrivani. Invited review - study design considerations. VRU 54.4 (2013).
Longitudinal: investigate changes over time. observations are made on each subject on more than one occasion. (information about event and outcome are collected at different times)
Cross sectional: individuals who are observed only once (information about event and outcome are collected during initial and only hospitalization.)
What is difference between event initiated and outcome initiated study designs?
Scrivani. Invited review - study design considerations. VRU 54.4 (2013).
Event intiiated - study designs were patients are selected by whether they have the event or not, and are then followed forard in time to determine if they develop the outcome. Cohort studies and clinical trials.
Outcome initiated - designs where patients are selected by whether they have the outcome or not, and then ar evaluated for possible associations to events that have occurred in the past. Case control studies.
