Clinical Research Flashcards
Pros of case studies
unique phenomena (it allows you to look at a rare phenomena); used to generate hypotheses-
Cons of case studies
Internal validity threats (e.g., lack of control- no systematic way to implement your control setting; no comparison group to compare) and external validity threats (e.g., hard to generalize it to the whole population)
Single Case Studies/Design
(A-B), or (ABAB):
Pros of single case
More internal control than case studies (e.g., standardized measure; systematic way you can ) and operationalized variables- defining what your variables are
Cons of single case
External validity threats (e.g., ); is removing a treatment ethical?
Correlational Designs
bc you’re not introducing any controls- it has the most vulnerability towards internal validity
Pros of correlational designs
Good external validity; descriptive; and generate hypothesis based on conceptual relationships
Cons of correlational designs
Internal validity threats (e.g., no random assignment or manipulation) and no causality
Pros of Experimental Design
random assignment!!!!!!!!!!
manipulation or control of variables
Cons of Experimental design
external validity threats (e.g., )
Threats to internal validity:
- history
- repeated testing-
- statistical regression
- selection bias
Threats to external validity
- sample characteristics
- stimulus characteristics
- reactivity
Sample characteristics
characteristics of research participants do not maponto population
E.g. ACT + ERP study with only PSYC 101 students (e.g., convenient sample- they are not going to reflect typical population with anxiety disorder, or OCD)
Stimulus characteristics
features of the study constrain generalizability:
E.g., highly trained therapists in RCT (often have clinicians- they are trained very well but in the real world, that is not the case because it is not implemented as well as highly trained therapists in RCT)
Reactivity
different response because know they are in a study/assessment
E.g., participants always complete H.W. for a DBT study
Statistical significance
difference scores (pre/post) on a measure are larger than chance; the difference between two groups’ scores on a measure is greater than chance
<5% chance the difference score would occur due to random error (e.g., internal/external validity threats)
Clinical significance
Did the intervention have a meaningful impact on function?
True/False: if something is statistically significant, it also means that it is clinically significant.
F
Reliable Change Index
did post-score move within 2 SD’s of mean score for non-distressed sample?
Psychotherapy historically guided by ___________
clinical experience and basic research
EBP originally a _______________ model
medical- very rigid, very black-or-white
Three factors in EBP (empirically based practice)
three factors: 1. clinical expertise 2. best research evidence 3. patient values and preferences
EBP leads to clinical practice guidelines
= somehow got translated to being EST (empirically supported treatment)
Publication bias
file-drawer problem;
highlighting less prominent, but less signficant results (misleading results) and withholding all studies that were not significant and reworking hypothesis
Challenges of EBP: presentation of statistical results
effect size (Bradley talked about this) empirically supported (vs. unsupported)
Challenges of EBP: proposed statistics
i. Percent recovered/percent improved (important what the denominator is- more liberal is when you look at percent recovered/people who completed the trial); e.g., starting with 100 participants; only 20 people completed and only 10 people got better- 10/20 = 50% sounds better)
ii. Residual post-treatment symptomotology: how are people functioning AFTER the treatment (they need to follow up with long-term treatement)
Western, Bradley article
- mimicking the medical model
- pitfalls of comparing to “no-treatment” or “intent-to-fail” treatments
- Exclusion criteria
Example of case studies
using ERP with dissociative identify disorder; CBT with child-onset schizophrenia (if it is rare then case studies)
Multiple baseline
treatment applied at different locations; time periods
Examples of correlational studies
Comparing the achievement test scores of individuals with Autism Spectrum Disorder who received early intervention to individuals with ASD who did not receive early intervention
Threats to internal validity: history
influence of other events (systematic influencing variable that is causing the change)
e.g. NCAA championship game: if we wnated to see depression medication: all groups seem to improve- so is that due to the medication or outside
Threats to internal validity: Repeated testing
familiarity and memory for responses e.g., achievement testing
Threats to internal validity: statistical regression
extreme scores move towards the mean; this is the concept that if you have extreme scores, the scores are going to be less extreme b/c they tend towards to average; if anything it’s going to come down or stay the same
e.g., post-tx testing of individuals with high anxiety
Threats to internal validity: selection bias
recruiting students in 8 am class for tx group and in 2 pm class for control group
Turner
Focused on publication bias
Presenting information in a descriptive mannner
Efficacy is to _______________ as effectiveness is to _________________.
Internal validity; external validity
Case studies take which of the following approaches to studying clinical phenomena?
Idiographic
When participants for a treatment outcome study are selected based on strict criteria and are randomly assigned to a control or treatment group they are likely participating in what kind of study?
Efficacy
Psychotherapy outcome studies that are interested in how well they work in the “real-world” are examining the therapy’s ______________.
Effectiveness
The Turner et al. study highlighted that the studies most likely to be published were ______________________.
the ones with significant and positive findings.
The authors contend that a pharmaceutical model for psychological treatment is not appropriate because many “treatments” are discovered _______________.
from clinical practice
In his TED Talk, Goldacre suggested that some scientists (not mass media) “rig” their data by _________________________.
giving extremely high or low doses of a treatment to participants in certain conditions.
EST studies make the assumption that a syndrome is discrete and can be treated via one modality. What is problematic with this assumption?
Most disorders co-occur/interact with symptoms of other disorders.
What is the purpose of a treatment outcome study?
is to determine how well their therapies work
What is the purpose of an analogue design?
When clinical psychologists seek to study clinical populations or situations but are unable to access them adequately, they may conduct a study with an analogue design.
Efficacy
how well it works in a lab; outcome for treated individuals is compared via clinical trial with outcome for individuals who receive alternate or no treatment
Participants in efficacy research
Participants are often selected on the basis of meeting strict (e.g., diagnostic) criteria and are randomly assigned to either a treatment group or a control group
Effectiveness
how well it works “in the real world,” or how well it translates from the lab to the clinics, agencies, hospitals, private practices, and other settings where clinical psychologists conduct therapy
