Study / Experiment Attributes Flashcards
(20 cards)
A research design is a plan for what?
For selecting and measuring the independent and dependent variables in order to answer research questions
Experimental Studies: those in which researchers ______ one of more factors that they will ________ during the experiment.
identify; manipulate or control
Nonexperiment Studies: those in which researches _______ existing conditions or differences without _______ them.
investigate; manipulating
Longitudinal Design/Research
an individual or group is followed for some amount of time (days or years)
Investigational Design aka “the natural experiment”
does not have any independent variables that are manipulated; one just observes what naturally happens.
Ex: survery of knowledge by SLP’s in area of ACC devices
Quasi experimental Design
involves the manipulation of at least one independent variable
ex: relationship of use of AAC devices in classrooms by children with cerebral palsy w/ educational advancement to grade level
continuous vs. categorical variables
Continuous variables=continuum of magnitude; ex: hearing loss in dB, MLU
categorical variables= no continuum– types of observations cannot be combined or scaled ; ex: measurement may be frequency counts within types of categories (nominal data)
What are advantages of single subject designs?
- close examination for individual differences
- a real subject will be described in data
*single subject designs does not necessarily mean one subject; can be a group
What are advantages of group designs?
- substantial number of statistical tools for assessing reliability of observations
- responses and existance of statistical controls over order, sequence and influence of independent variables.
single group design vs. group design
single group design : typical of individual & no assumptions of uniformity :: group design : typical of a group & assumption of uniformity
ex post facto design
(single or group) restrospective, but historical variables compared to finding effect of an independent variable
experimental design
involves the manipulation of at least one variable and the random assignment of subjects to groups
time series design
(single or group) A-B design (no treatment)
counterbalanced design
(group only) equal proportions of independent variables
randomized control trials
(group only) random assignment of subjects to groups; one group has placebo or conventional treatment & one or more groups have experimental treatment; blinded or double blinded
factorial design
(group only) subject organized in groups by independent variables; may have factors nested under factors
between subjects
(group only) measuring for a difference between 2 or more groups of subjects
within subjects
measuring whether a difference has occurred in one group over time (time series)
prospective vs. retrospective research
prospective: quetions formed before data collection:: retrospective: data existed before research
selection variables vs. treatment variables
selection: cannot be manipulated :: treatment: can be manipulated
