Experimental designs Flashcards
What are the different types of experimental designs:
Descriptive –Naturalistic observation, case study
Correlation – Case control study, survey
Experimental – Factorial design, Randomization.
DESCRIPTIVE DESIGNS:
What is a naturalistic observation?
Observations conducted in real-world situations (opposite of laboratory observation)
Must be specific: who, what, when & where is observed?
Time-interval sampling: e.g. first 5 minutes of each 30 minute interval
Event sampling: observed every time a specific event takes place - e.g. the frequency of aggressive acts such as hits, shouts and pushes observed in a playground?
DESCRIPTIVE DESIGNS:
- Case study
- Single case design
Case study:
-Detailed presentation of a single case, of a particular phenomenon,
-Case studies can generate hypotheses
-Case studies are important in rare diseases
-more then one case study = case series
Single case design:
It is easier to demonstrate causality in single-case design as the effect can simply be turned on and off.
ABA Design –>
A) Baseline: How frequent are violent episodes?
B) Give drug: Is violence less likely?
A) Withhold drug: Does the frequency of violent episodes increase?
B) Give drug again: Does it decrease again?
CORRELATION DESIGNS:
- Case-control study
- Survey
Case-control study:
-Compares the differences between people
-Observational and often retrospective (researching something that happened in the past).
-“Good for looking at trends in the population if you have historic data”
Survey:
-Sometimes the only option (clinical/ethical reasons)
-Sometimes the best option – assessing prevalence, incidence etc in a population
-Most useful when exploring concepts or relationships between variables
What are the characteristics of an experimental design?
-manipulation of the independent variable. The other one was control of covariate variables:
-The fact that we are comparing the results from one group with that of another to make sure the change was not due to another reason (e.g., gender or age). If all else is equal – the effect will likely be due to the independent variable.
-To compare a control and experimental group, we need to make sure that they are equivalent (overall, the same on all characteristics).
-One way to ensure that the two groups are as similar as possible is randomization in assigning people to groups
Define and explain:
-Factorial design
-Randomization
Factorial design:
-Sometimes, the effect of one variable is affected by the presence of another one. This is known as interaction, as is an important part of factorial designs.
-For example: the type of medication leads to differences in effect in comorbid disorders
-Baseline measure of dependent variable for both groups BEFORE and AFTER we introduce the independent variable
Randomization:
-When your groups are large enough, the chances increase that they are similar (e.g. age) and we need to make sure that:
a) they don’t affect the group differentially by spreading them across the group equally
b) measuring them and then accounting for their effect statistically
What is a randomized control trial?
What is it’s structure?
Randomized Controlled Trial (RCT)
-Investigator randomly assigns one group to an intervention, the other group gets nothing (placebo) or else an alternative intervention.
-Considered as very convincing evidence of cause-and-effect relationship
Structure:
-Randomly allocate participants to either:
- Experimental (or ‘treatment’) group - those who get the independent variable (e.g. caffeinated coffee)
- Control group = Those who don’t get the independent variable (e.g. decaffeinated coffee)
Compare the dependent variable of the two groups.
What are the STRENGTHS AND LIMITATIONS of a randomized control trial?
Strengths:
-Having a control group means you are able to find out exactly what effect the IV has on the DV
-If designed well, RCTs are robust against alternative explanations
-Results are easy to interpret
-Design minimizes researcher subjectivity
Limitations:
-Usually requires a large number of participants - Informs us about averages but not the individual - Expensive
-Controlled settings, are the findings valid in an applied setting? (i.e., external validity)
-Limited outcomes possible, cannot change during the study process - What is the control? - Placebo psychotherapy?
What is a non-randomized controlled trial?
Sometimes, we can’t randomly select and we determine group allocation before the trial
Experimental group
- who decided themselves that they will get the independent variables (e.g. medication)
Control group
- who decided themselves that they won’t get the independent variables (e.g. placebo)
Apart from lack of randomization of groups, everything else can be the same as in the randomized control trial. We still compare the two groups
What are quasi-experimental designs?
-Developed to provide alternative means for “examining” causality in situations not conducive to true-experimental designs, such as ethical issues (e.g. withholding treatment to control group)
-Allows some examination of causality in situations where complete control is not possible
At least one of the following three elements of true-experimental research is missing:
- Random sampling is not possible (most common)
- There is no control groups
- Manipulation of the treatment (independent variable)
FACTORS THE INFLUENCE CHOOSING A DESIGN:
-History effects
-Maturation effects
-Testing effects
History effects: Any external event that occurred between your pre-test and post-test that caused the change in measure, and not the intervention (e.g., COVID-19)
Maturation effects: Change simply with time and not due to exposure to the treatment / intervention . For example, individual recovery time from stroke or traumatic brain injury
Testing effects:; The fact that people were tested, not the treatment, caused the change in behavior.
FACTORS THE INFLUENCE CHOOSING A DESIGN:
-Instrument decay
-Attrition
-Regression to the mean.
Instrument decay: Gradually loss of accuracy of measurement. Usually a problem when using frequency data (e.g. recording number of cigarettes smoked). Effect not due to intervention but the fact that fewer events were recorded.
Attrition: Loss of subjects during the study, Remaining subjects might not be representative
Regression to the mean: Measurements often vary in a zig-zag fashion When you identify your sample by a measure that happened to be on a zig, a zag will follow even when your intervention had no effect at all!
