Psych R&D

• Main effect 


  overall significant effect of a particular independent variable collapsed across the levels of other variables
• All factorials need…


  more than one IV or GV
• Interaction 


  The effect of one IV (factor) depends on the level of the other IV (factor)
• # of main effects in a design

  Can be as many main effects as there are factors

  • 2 x 2 = 2 main effects
  • 2 x 3 = 2 main effects
  • 2 x 3 x 2 = 3 main effects
• Confounds


  Extraneous variables that vary in some non-random way across conditions
• Sequences


  type of confound where people change as they spend time working on tasks
• Carryover


  Type of confound where responses to one condition influences the response to a later condition
• Which three factors statistically influence the likelihood of findings statistically significant differences between two sample means?

  • Variance/error
  • effect size
  • sample size
• role of effect size on significance 

  The average difference between your conditions (effect size) determines the numerator of your t or F ratio; The greater the average differences between your conditions, the higher your tor F obtained, thus increasing the chance it will be significant.
• role of variance/error on significance

  Standard deviation (i.e., amount of variance or error in your data set) determines the denominator of your t or F ratio; The less variance you have, then the lower the denominator of your tor F ratio will be; Lower denominators will result in higher t or F obtained, thus increasing the chance it will be significant.