Notes on Studies Comparing Two Means

  • Introduction Theory

    • Eating behavior in anorexic individuals is linked to personal control.
    • Anorexics often exert control over eating when they feel a lack of control in their lives.

  • Sample

    • 12 female participants with high anorexic tendencies assigned to two groups:
    • High control group
    • Low control group

  • Hypothesis

    • High control participants will consume more sweetened breakfast cereals compared to low control participants.
    • Expected: Mean number of cereal pieces eaten will be higher in the High control group.

  • Design

    • Randomized two-group, between-subjects design:
    • IV: Level of personal control (High vs. Low)
    • DV: Number of cereal pieces eaten

  • Experimental Hypothesis

    • Non-zero difference between population means (effect size).
    • High control group expected to have a larger population mean of cereal consumption.

  • Sampling Error Effects

    • Overestimate: Sample mean may inaccurately reflect a larger difference than population means.
    • Underestimate: Sample mean may reflect a smaller difference than actual population means.
    • Wrong Direction: Sample means can differ opposite to the true population difference.

  • Null Hypothesis

    • Assumes no difference between population means (i.e., High control mean = Low control mean).
    • A significant difference (p < .05) leads to rejection of the null hypothesis.

  • t-tests

    • Independent-samples t-test: Compare means of two independent samples.
    • Paired-samples t-test: Compare means of two related samples.
    • One-sample t-test: Compare sample mean to a known mean.