Research Appraisal - Eppp Skills

1. Identify the best generalizability statement: RCT excludes comorbidities; what is reduced?

   Ecological validity (generalizability to real-world clinical cases). Excluding complex cases reduces the study's relevance to everyday clinical scenarios, limiting real-world applicability.

2. Identify the main consequence of violating independence due to clustering.

   Standard errors are underestimated; Type I error inflates. Violating independence through clustering leads to correlated errors, biasing variance estimates and increasing false positives.

3. Identify the primary threat: groups differ at baseline because participants self-select treatment.

   Selection bias (confounding by indication). Self-selection introduces confounding, as baseline differences may explain outcomes rather than the treatment itself.

4. What is random assignment, and what does it primarily protect against?

   Assigning by chance; reduces selection bias and confounding. Random assignment equalizes groups on potential confounders, enhancing internal validity by preventing systematic differences.

5. Which option best defines selection bias in nonrandomized studies?

   Systematic baseline differences between groups before treatment. Selection bias undermines internal validity in nonrandomized designs by creating pre-existing group differences that mimic treatment effects.

6. What is construct validity for a measure or manipulation?

   Degree to which it accurately represents the intended construct. Construct validity evaluates if the operationalization faithfully captures the theoretical concept it aims to measure or manipulate.

7. What is attrition bias (differential dropout) in longitudinal or RCT research?

   Unequal loss to follow-up that distorts group comparability. Attrition bias compromises group equivalence over time, potentially biasing results toward remaining participants' characteristics.

8. What is a Type II error in hypothesis testing?

   Failing to reject a false $H_0$ (false negative). Type II error happens when a real effect is missed by not rejecting the null hypothesis.

9. What is external validity in research appraisal?

   Generalizability of findings across people, settings, and time. External validity assesses how well study results apply beyond the sample to broader populations, contexts, and periods.

10. What is experimenter expectancy (Rosenthal) effect?

    Researcher expectations unintentionally influence participant outcomes. Experimenter expectancy biases results through subtle cues that influence participants toward expected behaviors.

11. Identify the correct interpretation of a $p$-value in null hypothesis testing.

    Probability of data (or more extreme) given $H_0$ is true. The $p$-value quantifies the likelihood of observing the data assuming the null hypothesis holds, aiding in significance decisions.

12. What is statistical power, expressed in terms of Type II error rate $\beta$?

    $1-\beta$. Statistical power represents the probability of correctly detecting a true effect, directly related to avoiding Type II errors.

13. Find the correct conclusion: convenience sample from one clinic; what validity is most limited?

    External validity. Convenience sampling restricts the applicability of findings to diverse populations, thus limiting generalizability.

14. What is demand characteristics as a threat to validity?

    Participants alter responses based on perceived study expectations. Demand characteristics compromise validity by cueing participants to behave in ways that confirm hypothesized outcomes.

15. What is a Type I error in hypothesis testing?

    Rejecting a true $H_0$ (false positive). Type I error occurs when the null hypothesis is incorrectly rejected, leading to erroneous claims of an effect.

16. What is statistical conclusion validity?

    Accuracy of inferences about covariation and statistical results. Statistical conclusion validity verifies the reliability of conclusions drawn from data analysis regarding relationships and significance.

17. Identify the strongest design for causal inference between an IV and DV.

    Randomized controlled experiment. This design minimizes biases through randomization and control, enabling strong causal claims about variable relationships.

18. Which option best describes the assumption of independence in common tests?

    Each observation is not influenced by or duplicated across others. The independence assumption ensures that data points are unrelated, which is crucial for valid statistical inference in parametric tests.

19. What is the Hawthorne effect as a methodological threat?

    Behavior changes due to awareness of being observed. The Hawthorne effect introduces reactivity bias, where observation alone alters behavior, threatening internal validity.

20. What is internal validity in research appraisal?

    Confidence that the IV caused the DV within the study. Internal validity ensures the study's design isolates the effect of the independent variable on the dependent variable, minimizing alternative explanations.

21. What is a confound in evaluating a study's methodology?

    A third variable that covaries with IV and could cause the DV. Confounds threaten internal validity by providing alternative explanations for observed effects between variables.

22. What is the primary purpose of blinding in clinical trials?

    Reduce expectancy and measurement bias in participants/raters. Blinding prevents bias by concealing group assignments, ensuring objective responses and assessments.

23. Choose the correct issue: many outcomes tested with no correction; what error risk increases?

    Type I error inflation (multiple comparisons problem). Without adjustments for multiple tests, the family-wise error rate rises, increasing the chance of spurious significant findings.

24. What is a placebo control intended to address methodologically?

    Nonspecific treatment effects (expectancy and attention). Placebo controls isolate active treatment effects by accounting for psychological and contextual influences on outcomes.

Study these flashcards