Experiments: Confounding and Obscuring Variables

20 vocabulary flashcards summarizing key terms related to confounding and obscuring variables, reasons for null effects, and solutions to improve experimental designs.

Design confound

An extraneous variable that varies systematically with the independent variable (IV), offering an alternative explanation for the results.

Selection effect

A bias that occurs when participants in one group differ from those in another group before the study begins.

Order effect

When earlier tasks influence performance on later tasks in a within-groups design (e.g., practice or fatigue).

Null effect

No statistically significant difference between groups on the dependent variable (DV); the IV appears to have no impact.

Weak manipulation

IV levels are too small or subtle to create meaningful differences in the DV.

Insensitive dependent variable

A DV measure that lacks sufficient resolution to detect subtle changes (e.g., pass/fail scoring).

Ceiling effect

Scores cluster at the high end of a scale, obscuring potential group differences.

Floor effect

Scores cluster at the low end of a scale, obscuring potential group differences.

Measurement error

Random noise introduced by unreliable or imprecise measurement tools.

Individual differences

Variability in participant traits (e.g., motivation, ability) that adds noise within groups.

Situation noise

Uncontrolled external distractions or environmental factors that increase within-group variability.

Between-group difference

Variation attributable to the IV across different experimental groups.

Within-groups variability

Variance among participants within the same group that can obscure true effects.

Within-groups design

Design in which each participant experiences all levels of the IV, reducing individual-difference noise but vulnerable to order effects.

Matched groups

Assignment method where participants are paired on relevant traits before being split into different conditions to control for individual differences.

Stronger IV manipulation

Increasing the intensity or distinctiveness of the IV conditions to produce detectable effects.

Sensitive DV

A precise, multi-level measurement capable of detecting subtle changes in the construct of interest.

Control environment

Standardizing the experimental setting to minimize situation noise and extraneous variability.

Increase sample size

Recruiting more participants to reduce random error and improve estimate precision.

Statistical power

The probability of detecting a true effect when it exists; boosted by larger samples, stronger manipulations, and reduced variability.