Exam 2 - Experimental settings Flashcards
What are 3 important factors for an experimental study design?
- has an Intervention of treatment
- study design controls extraneous variables (including control group)
- Randomized – not always, but usually
Define randomized control trials (RCT)
- Research method that uses random assignment of participants to compare the effects of different treatments of interventions
- Gold standard for evaluating the effect of an intervention, treatment, or program
- Utilizes efficacy studies (ideal conditions) rather than effectiveness studies (real-world conditions)
- Intervention group vs a placebo group (no or low intervention)
Describe the two RCT study designs
1) Parallel group studies
- Screened for eligibility criteria and consented to study
- Randomly assigned to intervention group or control group
- Outcomes are recorded and assessed
- Called parallel because both interventions are occurring in parallel (at the same time)
2) Cross over studies
- Participants will be both the intervention and their own control which reduces variability
- Target population identified, screened, and consented
- Randomized to intervention (supplement or dietary component of interest) or placebo for “x” amount of time
- Washout period occurs to normalize whatever the intervention was
- Then switch groups so intervention group now in control period, and control now in intervention period
Distinguish the difference between clinical RCR and controlled clinical trials.
Clinical RCR
- Test new treatments, drugs, or medical devices with human participants to assess efficacy and safety
Controlled clinical trials
- Does not randomly assign to intervention or control group
- Population being studied may be unique and not able to be randomized
- Quasi experimental study (not randomized)
What are the 9 steps in planning a dietary intervention?
1) Define hypothesis and main objective
2) Selection of primary and secondary outcome measures
3) Define study population
4) Selection of study design
5) Definition of measurements
6) Development of the study protocol
7) Estimation of study costs
8) Ethical approval and study registration
9) Implementation of the study
Define outcome.
- Quantitative measure related to the health condition or disease under investigation
- Primary outcome: main outcome associated with the study (ex: cancer)
- Secondary outcome: used to evaluate additional effects of the intervention, shed light and provide more information on primary outcome (ex: death from cancer, biomarkers)
What types of outcomes are typical for an intervention study?
- Continuous: weight, biochemical parameters
- Binary: yes or no events
- Event time: survival time
- Count: frequency of occurrence of event
- Categories: absent, mild, moderate, or severe
What are control groups? What types of control groups can be in an intervention study?
- Control groups: those receiving no or low interventions in comparison to the intervention groups
What factors should be considered in selection of study participants?
- Eligibility criteria: age, gender, health and disease status, demographic variables, lifestyle factors
- Thoroughly define the study population
- Target population for the test intervention
- Should be specific on who participants are (dependent on what the study is about)
What does inclusion and exclusion criteria mean in an intervention study?
Inclusion
- Criteria that are essential for potential subjects to be eligible to participate
- Remove the influence of specific confounding variables
- Target population
- Ex: age, race, ethnicity, gender, risk for certain diseases, tobacco use, etc
Exclusion criteria
- Criteria that will exclude potential subjects from participating in the study or require their removal as subjects
- Confounders
- Ex: history of prior illness, failure to adhere to pre-test requirements, conditions that make them unable to complete required testing
What is the difference between type I and type II error?
Type I error
- “False positive”
- Detects a difference when no difference exists
- Ex: telling a male they are pregnant
Type II error
- “false negative”
- Fail to detect a difference when there is a difference
- Ex: telling pregnant woman she isn’t pregnant when it’s obvious she is
Define power. What is the relationship between power and type II error?
Power: the probability that a clinical trial will reject the null hypothesis (p < 0.05) – difference between the study groups is statistically significant
- At least 80% is desirable, 80% confident that the observations or difference between the group is not due to chance
- Power increases as sample size increases
- Power increases as the chances of committing a type II error decrease (false negatives)
Define effect size. What is the relationship between effect size and sample size?
Effect size: difference between the value of the variable in the control group and that In the intervention group; how much better do you expect treatment group to be
- Small effect size = large sample size needed
- Large effect size = small sample size needed
- Too small sample size could fail to detect differences, too large sample size could waste resources
Variance: how much variability do you expect in measurement
- Large variation = larger sample size required
Briefly explain why randomization is important in a study
- Use of chance mechanisms to assign treatments
- Important to avoid selection bias; people may act or react differently if they know they have a specific intervention
- Best assurance that control group is like the intervention group; only way to control for known and unknowns
What is a quasi-experimental study?
- Studies where there is not a control group, or where it’s unethical to have a control group that does not receive the intervention (ex: vitamins to infants, cancer treatment)
- Easier, quicker, cheaper
- Lacks random assignment or missing control group
