Module 12 - Evaluating Research Flashcards
(27 cards)
What does a p-value indicate?
A probability value indicating how likely results are due to chance. A p-value below 0.05 is traditionally considered ‘statistically significant’.
The p-value helps researchers determine the significance of their results in hypothesis testing.
Define P-hacking.
The misuse of data analysis to obtain significant p-values, often through selective reporting or trying multiple statistical tests.
P-hacking can lead to misleading conclusions and a lack of reproducibility in research findings.
What is the File-drawer problem?
The tendency for non-significant or negative research results to remain unpublished, skewing the available evidence.
This problem can lead to a biased understanding of research areas, as only positive results are often shared.
What is Prior Probability?
The likelihood that a hypothesis is true before conducting a study. Affects how we interpret significant results.
Understanding prior probability can help in evaluating the strength and relevance of new findings.
What does the Reproducibility Crisis refer to?
A large number of published studies (especially in psychology) cannot be replicated, raising questions about their validity.
This crisis highlights the importance of robust research methodologies.
What is Replication in research?
Repeating a study to see if the original results can be obtained again under similar conditions.
Replication is essential for confirming the reliability of research findings.
What is Preregistration?
The act of publicly recording a study’s methods and hypotheses before data collection to prevent selective reporting.
Preregistration enhances transparency and credibility in research.
Define Internal Validity.
The extent to which a study shows that the intervention (independent variable) caused the outcome (dependent variable).
High internal validity means that the study’s design effectively controls for confounding variables.
What does External Validity refer to?
The degree to which study findings can be generalised to other people, settings, or times.
External validity is crucial for the applicability of research findings in real-world contexts.
What is Empirical Evidence?
Data obtained through observation or experimentation rather than theory or opinion.
Empirical evidence is foundational in scientific research for establishing facts.
List threats to Internal Validity.
- History: Events outside the study
- Maturation: Participant changes over time
- Testing: Repeated measurement effects
- Instrumentation: Changes in measurement methods
- Regression to the Mean: Extreme scores returning to the average
- Selection Bias: Non-random sample differences
- Attrition: Dropout of participants affecting results
- Diffusion of Treatment: Control group exposure to treatment
Each of these factors can lead to incorrect conclusions about causality.
Define Efficacy in the context of treatments.
Whether an intervention works under controlled conditions (e.g., clinical trials).
Efficacy is often assessed through rigorous experimental designs.
What is Effectiveness?
Whether an intervention works in real-world settings (e.g., schools, homes).
Effectiveness assessments consider practical implementation and outcomes.
What does Efficiency refer to in evaluating treatments?
The cost-effectiveness and accessibility of a treatment.
Efficiency is important for determining the feasibility of interventions in various contexts.
What is Evidence-Based Practice (EBP)?
An approach that integrates the best research evidence, clinical expertise, and patient values.
EBP aims to improve healthcare outcomes by using the best available evidence.
What is a Randomised Controlled Trial (RCT)?
A study where participants are randomly assigned to different groups to test cause-and-effect relationships.
RCTs are considered the gold standard in clinical research.
Define Small-N / Single-Case Design.
A study design focusing on a small number of participants, where each acts as their own control.
This design is useful for in-depth analysis of individual cases.
What does Efficacious mean regarding treatment levels?
Backed by multiple well-designed, randomised studies showing consistent results.
Efficacious treatments are considered reliable and effective.
What does Probably Efficacious mean?
Supported by strong but less rigorous or fewer studies.
This indicates some evidence of effectiveness, but not as robust as ‘efficacious’.
What does Possibly Efficacious indicate?
Limited support, typically from a single study without randomised controls.
Such treatments may require further investigation to confirm their effectiveness.
What does Not Empirically Supported mean?
Based only on anecdotal or unverified reports.
These treatments lack scientific validation and should be approached with caution.
What does Discredited mean in treatment efficacy?
Proven to be ineffective or harmful.
Discredited treatments should not be used in practice due to their negative outcomes.
What is Basic Research?
Theory-driven, curiosity-based exploration without direct application.
Basic research is fundamental for generating new knowledge and theories.
What does Translation to Humans involve?
Testing basic research findings with small human samples (pre-clinical trials).
This step bridges the gap between laboratory research and clinical application.
