Quantitative Study Designs and Sampling Flashcards
What are the five questions to guide main research elements in a study design?
1) Are the people in the study assigned to groups?
2) How many measurements are being used?
3) What types of measures or observations are being used?
4) Is there an interest in generalizing the findings to other populations or settings?
5) Can you conclude that the findings are based on the manipulation of the independent variable?
Research Synthesis (Meta-analysis)
A systematic quantification of the results across multiple studies
Surveys
A descriptive research method generally broad in scope.
Methods often include questionnaires, normative, surveys.
Descriptive Research Methods
Developmental research (longitudinal, cross-sectional)
Case study
Observational research
Correlational research (explores relationships among variables)
Epidemiology
Studies the distribution and determinants of physical activity and health-related states in populations
Experimental Designs
Pre-experimental, quasi-experimental, and true experimental design.
Attempt to establish cause and effect relationships.
Includes the manipulation of an independent variable to see the effects on a dependent variable
What conditions are necessary to make a causal claim?
Causes need to precede effects in time.
There needs to be a correlation between causes and effect.
The relationship between the causes and effects can’t be explained by other variables.
What is a pre-experimental study design?
Used when a true experiment and quasi-experimental designs are not possible (used as prequel or precursor to a randomized control trial).
Combination of pre-test, post-test testing and instruction.
Used when resources are low.
Fail to include at least one or more of the following…
-a pre-test, control group, a comparison group, and/or randomization to groups
Pose challenges to internal and external validity since many of the threats are not controlled in this type of study design.
May be less costly both financially and in terms of time, as well as potentially imposing less burden on participants.
Should be used with great caution because of the challenges it poses for researchers hoping to identify causal relationships among variables.
What are the different designs of a pre-experimental study?
One-time case study: participants are exposed to an intervention and then are assessed on the outcome of interest.
One-group pre-test-post-test design: participants complete an assessment, are exposed to an intervention, and then complete an assessment again.
Post-test-only with non-equivalent groups design: this design is used after an intervention is implemented with one of two intact groups; both the experimental group and the comparison group complete a post-test assessment.
Alternative treatment post-test-only with non-equivalent groups design.
What is a true experimental study design?
Most powerful means of generating new knowledge.
Only quantitative design that can be confidently used to identify a cause-and-effect relationship. Conducted in lab within a controlled environment.
Level of control helps to ensure study’s internal validity
-any changes in an outcome is the result of a treatment or intervention and not a result of other factors related to the sample, the measures, the techniques, and other possible potential threats.
Not as easy to claim external validity given the trade-off needed to ensure internal validity
-potential application of the results of the study or generalizability
Independent variable manipulation (ie. treatment).
Control and experimental groups (has at least one experimental group (exposed to manipulation) and one control group (members not exposed to the manipulation)).
Random selection ideal, but not always possible.
Random assignment used to determine who is in what group
-each person in sample as equal chance of ending up in either group
-ensures groups are equivalent prior to the study manipulation
Theory driven
-goal is to allow researchers to collect data and test their hypothesis with minimum influence or contamination from extraneous variables
-can be used to claim cause and effect relationship whereby it is logically determined that one factor has a predictable influence on another (extraneous variables need to be controlled)
What are the different designs of a true experimental study?
Pre-test-post-test control-group design: used to examine change in the dependent variable that can be attributed to the independent variable; want to see which group changes more as a result of the manipulation or treatment.
Post-test only control-group design: the independent variable is introduced to the randomly assigned experimental group participants, and then effects of the treatment vs no treatment are tested by looking at whether the groups are different.
Solomon four-group design: groups experience different tests/different order of tests/tests occur at different time points for each group.
What is a quasi-experimental study design?
Random assignment of participants to groups has not been accomplished.
Approximate the conditions of the true experiment but do not allow for the control and/or manipulation of all relevant variables.
A researcher must clearly understand the compromises between internal and external validity
-low internal validity because groups are not likely to be equivalent
-difficult to make claims that any changes seen between the pre and post-test are due only to the manipulation of the IV
-higher on external validity because people studied are in natural environment, making generalizability more feasible.
People studied in real world settings, an IV is introduced/manipulated, and there is a DV that is measured
What are the different designs of a quasi-experimental study?
Non-equivalent (pre-test and post-test) control-group design: one group experiences manipulation and other group is the control.
Single-group interrupted time-series design: one group experiences the manipulation by interrupting the control
Control-group interrupted time-series design: on group experiences manipulation interrupting control then going back, other group only experiences control.
What are single subject designs?
A-B-A single subject design: single person or small number of people are studied over a long period of time
-naturally occurring behaviour of an individual is examined over time
-baseline behaviour is assessed, treatment provided, and then the treatment is withdrawn and the behaviour is assessed again over a period of time
Population
All the individuals of interest to the research.
The larger group from which a sample is taken.
