Week 3 Flashcards
What are common names for experimental research designs?
Randomized trial
Clinical trial
Randomized clinical trial
Randomized controlled trial
Extraneous variables
Any factor not related to the purpose of the study that could affect the dependent variable (age)
When does an extraneous variable become a confounding variable?
When they are not controlled
Why is it important to randomly assign subject?
Subject judgement does not play a role
Gives an equal chance of being assigned to each group
Helps control extraneous variable
Concealed allocation
The person deciding if someone will be included in the study has not seen what group the subject is in
Random sample
Method of recruiting a sample from a certain population
Random assignment
Used to equalize two groups (treatment and control)
Designs for independent groups between subject designs
Examining the differences between independent groups of studies (groups get different things)
Repeated measures design (within subject design)
One group measured multiple times
Order effects
Outcome may be dependent on the order of the measurements (randomized order)
Crossover design
Single factor pretest and posttest
Washout period
Period of time to eliminate the prolonged effects of the treatment
Within subjects multifactor design
Same group with do a squat with weights and then without
Between subjects multifactor design
One group with do squats with weights and another group does squats without weight
Mixed model design
1 independent variable is between subjects and 1 independent variable is within subjects
Random sampling
All items have the same chance of selection (minimizes bias)
Systemic sampling
Potential subjects are organized according to an identifier like medical record number
What are types of probabilistic sampling?
Random and systemic
What are types of non-probabilistic sampling?
Convenience
Snowball
Purposive
Convenience sampling
Recruit easily available people who meet criteria of study (most commonly used and is volunteer)
Snowball sampling
Start with a few subjects and have them tell others
Purposive sampling
Make specific choices about who will serve as subjects by handpicking
Responsiveness
Ability of an instrument to change over time
Why do we measure change?
To see if patient has improved or declined
Minimal detectable change (MDC)
Ability of an instrument to detect change beyond measurement error
Minimal clinically important difference (MCID)
Ability of an instrument to detect minimally important change
Bias
Misleading results
Random error
Unreliable results
What are acceptable percentages to lose in a study?
<5% is good
>20% is bad
Per protocol analysis
Remove them from study
Intention to treat
Keep all of participants data in the group they were randomly assigned to
Why is intention to treat better than per protocol?
Manages bias the best