L2: Interventions and RCT's Flashcards
What is a randomized control trial? (RCT)
EX?
An experimental form of impact evaluation in which the population receiving the program or intervention is randomly chosen from an eligible population
EX: Family Nurture Intervention in the NICU (Hane et al., 2015)
What are strengths of RCT’s?
Only type of study to establish causation
Able to assign & administer treatment in controlled way
Decreases selection bias
What are the weaknesses of RCT’s?
no pre-test - assumption all ppts. are the same (hence intervention must be the cause)
High cost
Logistically challenging
Results may not mimic real life treatment due to inclusion/exclusion criteria
ethics - remove treatment at the end of exp
generalisability of results
What is an experimental design?
EX
Systematic approach to research where researcher uses manipulation & control to understand casual processes
EX: Bandura et al., 1961: BoBo Doll Study
What are the limitations to identify causality in dev. psych?
Measuring from real life situations which may have negative outcomes for ppts
Difficult to measure everything in one person’s life
Majority of psych research aims to improve people’s lives
What is a longitudinal design?
EX
Researchers repeatedly examine the same individuals to detect changes which may occur over time
Aim to identify methods which correct or prevent problems in some developmentally vulnerable population
EX: Foley et al., (2022) Expectant mothers not fathers predict mind-mindedness
What are strengths of longitudinal designs
Ppt variables DO NOT affect data
Best way to spot developmental trends
Able to identify change and stability over time
What are the weaknesses of longitudinal designs?
Costly
Time consuming
Withdrawal; means if ppts who are left in the style all share a similar trait it creates bias
Data analysis varies in strength
What is a within subjects?
All participants are involved in manipulation
What is a between subjects design?
Half participants experience no manipulation, half do
What are strengths of within subjects?
Require less participants
Lack of individual variation
More powerful
What are weaknesses of within subjects?
Maturation - pre-test can prune ppts. (learning changes behaviour, not intervention?)
selection bias
Carryover effects/time elapsed: first test may influence how ppts behave on the 2nd test (perform better over time?)
What are the strengths of between subject’s design?
Shorter
Easier to set up
Avoid carryover effects
Minimise random noise
control group - compare outcomes
natural intervention
What are the weaknesses of between subject’s design?
Require more ppts
Complex
Individual variation of ppts
Placebo effect
selection bias
compensatory rivalry/resentful demoralisation - ppts. may give up if not in the intervention group
What did Costello et al., (2003) measure & find using a quasi longitudinal design?
Tested the role of social selection & causation of childhood psychopathy
After Casino opening to increase income; the ex-poor group fell to those of the never-poor children group, while levels among those who were persistently poor remained high
can’t be ascribed to family characteristics
What are the benefits & drawbacks of Costello’s quasi longitudinal design?
Benefits:
Natural Intervention
Use of standardized measures
Groups were physically separated
Large controlled sample
Drawbacks:
Placebo effect?
had to take resources away from ppts. at the end (financial support) - unethical
What is compensatory rivalry?
Members of the control group try to compensate for the fact that they are missing out on the experiment treatment by working/trying/concentrating harder.
What is resentful demoralisation?
Ppts may become resentful & demoralised because they haven’t received intervention - give up if placed in control group affecting results
EX. of a RCT: Bucharest Early Intervention Project (BEIP)
Examined the effects of institutionalized Romanian Orphans
Wanted to determine if these effects can be remediated e.g. through foster care
What did they study in the BEIP?
Recruited 136 kids from institutions
Randomly assigned to foster care to remain in institutionalized
Two Samples: Institutionalized vs non (From Romania)
Measured: language, cognitive, physical & social development, emotional reactivity, MH problems, genetics
What did the BEIP find?
BEIP research had promising results and demonstrated that placement in foster and family care resulted in significant improvements in the aforementioned areas, with the age at which a child was placed in family care being a significant predictor of outcome — the younger, the better
What should you consider when undergoing intervention?
Is the design appropriate?
Is there sensitive context?
Are there subjective beliefs involved?
Selecting a control group: awareness of group membership
Intent to treat
There may be unintended alterations to randomization such as ppts seeking out other interventions “bleed out effect”
What ethical considerations should be made?
When intervention should end
Equipoise: Professional disagreement between scientists & Drs. about what treatment is best
What should you consider when developing and evaluating interventions?
Is it feasible/acceptable?
Will ppts be responsive?
How close is the intervention to the original design?
How effective are interventions?
