Laboratory Experiment — Controlled Experimental Design

Both are true experiments with random assignment and manipulation of an independent variable. The distinction is setting: a laboratory experiment is conducted in a controlled, artificial environment designed by the researcher, while a field experiment is conducted in the participant's natural context. Lab experiments offer superior internal validity; field experiments offer superior ecological validity. The choice depends on whether the priority is causal precision or real-world generalizability.

Sample size depends on the expected effect size, the desired statistical power (typically .80), and the significance threshold (typically alpha = .05). For a medium effect size (Cohen's d = 0.50) in a two-group comparison, approximately 51 participants per group are needed. Use G*Power or a similar tool to conduct an a priori power analysis specific to your design before data collection begins.

Strictly speaking, no. Without a control condition or baseline, you cannot rule out the possibility that any change in the dependent variable is due to maturation, history, or regression to the mean rather than your manipulation. A control group — ideally a no-treatment or placebo condition — is essential to isolate the effect of the independent variable.

Single-blind procedures (participants unaware of condition) reduce demand characteristics and social desirability bias. Double-blind procedures (both participants and data collectors unaware of condition) additionally prevent experimenter expectancy effects — subtle cues a researcher might unintentionally give that influence participant behavior. Double-blinding is especially important when the dependent variable involves subjective judgment or behavioral ratings rather than objective automated measurement.

Construct validity refers to whether the operationalized variables actually measure the theoretical constructs they purport to represent. High internal validity confirms that the manipulation caused a change in the measured variable; construct validity confirms that the measured variable is a genuine instance of the theoretical concept. A laboratory experiment can be internally valid but construct-invalid if, for example, 'anxiety' is operationalized by a measure that primarily captures arousal rather than the cognitive appraisal component of anxiety.