Chapter 4 Flashcards - Quantitative Study Designs
What is the foundation for the design of the quantitative study?
Purpose statement
What is the study design supposed to show?
Methods of measurement, assessment and the statistical analysis
Quantitative designs can be used to address descriptive, predictive, and explanatory research problems
What type of study is the gold standard that helps understand cause and effect relationships?
Clinical Trials/Intervention Studies
What is validity in research?
How confident are we about the conclusion made as a result of the study we conducted
What is logical validity?
Refers to the quality of the researcher’s arguments, theory supporting needs of the study and appropriate interpretation of results based on the data.
I.e. If a study was conducted might be asked if you provided convincing justification for all the actions and outputs of your study
Essentially, if the researcher has recorded everything correctly, then the results should be right
What is construct validity?
Whether the measures used by researchers, test what they intended to measure (the idea/concept)
I.e. Measuring a child’s height on a growth chart mounted would likely give you the height of the child - the measures used accurately to get a proper test result
What are the five questions asked to determine what type of experimental design a study may be conducted as?
- Are the people in the study assigned to groups?
- How many measurements are being used?
- What types of measures or observations are being used?
- Is there an interest in generalizing the findings to other populations or settings?
- Can you conclude that the findings are based on the manipulation of the independent variable
If the answer is “No” to the question “Are people in your study assigned to groups?” then what type of experimental design could that be and why?
A non-experimental or Pre-experimental study design and this is because there is no other group to compare or there are two groups but they are not randomly assigned
Non-experimental means there is no manipulation of any variables or interventions that are in mind. Hence it’s not really experimental
What is the problem with pre-experimental study designs?
No pre-test, no control group meaning no comparison group either and finally no randomization to groups
Saying Yes to the “Are people in your study assigned to groups?” and Yes to “Are people randomly assigned to groups?” what type of experimental design is it?
True experiment; study with two or more groups, individuals are assigned randomly to groups (control and experimental)
What would indicate a quasi-experimental design?
Having two or more groups but no random assignment to control and experimental
Essentially it’s studies that evaluate an intervention but do not use randomization groups
What is internal validity?
The researcher’s ability to claim that any change in the outcome is the result of a treatment or intervention and not the results of factors related to the measures, techniques or any threats
Basically a cause (study design) and effect (outcome) relationship was established in the research and no other factors actually affected it like selection bias
What is external validity?
Examining whether the study findings can actually be applied to other studies (generalized) but it would be difficult with internal validity being in a tightly controlled environment so it can’t be reproduced elsewhere
A true experiment has two things which are?
Experimental group (treatment or intervention group) & control group (not exposed to manipulation)
Textbook states: Any study that is designed with random assignment of participants, a control group, and manipulation of the independdnt variable is considered a true experiment
What are extraneous variables?
Variables that are not the independent but can still change the outcome of the dependent
They are commonly considered control variables
What type of design is shown here?
R = random assignment, O = observation or measurement, X = treatment
- The top is a pre-test post-test design in which you have the participants at a baseline, look at them before the intervention, make them go through the intervention and observe the result of the intervention on them after
- Bottom is a post test only where you have no pre-test; only measure the affect of the experiment on one group and then compare the two groups at the end
Explain how this is a repeated measures design
You assign a person to one of the three variables (Everyone get’s assigned to a independent variable) & you manipulate their diet every week to see what the dependent variable turns out as
I.e. Participant 1 gets assigned to every independent variable that participant 2 and 3 are assigned to but they start off in a different order for the variable, same with the others
Explain what is happening in each quasi-experimental study?
- Non-equivalent is just observing a pre-test of two groups, one is the control and the other is the experimental and you administer the treatment and look at the observed results after
- Interrupted time is observing the participants, then interrupting the time and putting the treatment in place and seeing the results of that. May need to register drug multiple times to get a couple of results
- Equivalent time sample design is where you see a pattern, where you administer drug and observe, then you administer a placebo and observe
What is a static group comparison?
A.K.A Post-test only with non-equivalent groups
Explain how this is an example of Pilot and Feasible Studies turning into a high-quality RCT
Did a double blind experiment where they studied male physicians whether or not taking asprin every other day would lead to less Heart Attacks and they say the aspirin group has less heart attacks but they had to account for diet, exercise, etc.
- Only able to do this because they had to make sure they could recruit people and whether or not they would adhere to the follow up questions (Done in the feasibility and pilot studies to allow for it to become a bigger RCT test)
