8.4: Individual Differences and Variability

• How do we prevent high variability from obscuring any treatment effects?

  The scores that are obtained in one condition are noticeably different than the scores in the second; thus, the absolute size of the difference must be evaluated in relation to the variance of the scores.
• Variance

  A measure of variability obtained by computing the average squared distance from the mean.
• Variance Within Treatments

  A measure of the differences between scores for a group of individuals who have all received the same treatment. The intent is to measure naturally occurring differences that have not been caused by a treatment effect.

  • Big differences between-treatments are good because they provide evidence of differential treatment effects.
  • Big differences within-treatments are bad because the differences that exist inside the treatment conditions determine the variance of the scores, large variances can obscure patterns in data.
• How do we minimise variance within treatments?

  1. Standardising procedures and treatment setting.
  2. Limit individual differences.
  3. Random assignment and matching.
  4. Sample size.