Which issue most directly threatens internal validity in a psychology experiment?

Which issue most directly threatens internal validity in a psychology experiment?

• A. Randomization
• B. Blinding
• C. Sample size
• D. Confounding variables

Answer: (D)

Internal validity is about ensuring that changes in the dependent variable are caused by the manipulation of the independent variable, not by other factors. Confounding variables are factors that vary with the manipulation or differ between groups in a way that could account for the observed effect. They introduce alternative explanations for the results, which directly undermines causal conclusions. For example, if participants in one condition receive more attention from the researcher, or if the experimental environment differs between groups, those differences could drive observed outcomes rather than the intended manipulation. That’s why confounding variables are the key threat here—they can mimic or mask the true effect of the manipulation. Randomization, blinding, and adequate sample size are important for reducing bias and increasing reliability, but they don’t inherently threaten internal validity the way confounds do. Randomization helps equalize unknown factors, blinding reduces expectancy effects, and larger samples improve power and precision. The central concern for internal validity remains the presence of confounding variables that offer alternate explanations for the results.

Internal validity is about ensuring that changes in the dependent variable are caused by the manipulation of the independent variable, not by other factors. Confounding variables are factors that vary with the manipulation or differ between groups in a way that could account for the observed effect. They introduce alternative explanations for the results, which directly undermines causal conclusions.

For example, if participants in one condition receive more attention from the researcher, or if the experimental environment differs between groups, those differences could drive observed outcomes rather than the intended manipulation. That’s why confounding variables are the key threat here—they can mimic or mask the true effect of the manipulation.

Randomization, blinding, and adequate sample size are important for reducing bias and increasing reliability, but they don’t inherently threaten internal validity the way confounds do. Randomization helps equalize unknown factors, blinding reduces expectancy effects, and larger samples improve power and precision. The central concern for internal validity remains the presence of confounding variables that offer alternate explanations for the results.