Tests of a difference week 5

• 1. Tests

  • A t-test examines the difference between two sets of scores
  • Emphasis on mean differences within a sample and whether this can be applied to the overall population
• Within Subjects


  • Participants experience all experimental conditions
  • Also known as related designs, dependent samples, repeated measures designs
• Between Subjects


  • Participants can only belong to one experimental group
  • Also known as unrelated designs, independent samples
• Within Subjects Experimental Designs

  • Participants testing before and after an intervention
  • Lap times after drinking different sports drinks
• Between Subjects Experimental Designs

  • Age group comparisons
  • Gender differences
  • Experimental vs. control group studies
• Participants in within subjects designs experience all experimental conditions, while participants in between subjects designs can only belong to one experimental group
• Experimental Designs

  • Within Subjects
  • Between Subjects
• Tests of a Difference

  • Parametric
  • Non-Parametric
• Effect size tells us about the magnitude of the differences between groups
• SPSS does not provide effect size data, but Eta squared (η2) can be calculated using the information in the output
• Calculating Eta Squared for a between groups design

  0.6% of the variance in the dependent variable is explained by the independent variable
• Calculating Eta Squared for a within groups design

  50% of the variance in the dependent variable is explained by the independent variable
• Statistical analysis


  Independent-samples t-test was conducted to compare the self-esteem scores for males and females
• Assumptions check


  Preliminary analyses were performed to ensure no violation of the assumptions of normality and homogeneity of variance
• To meet the assumption of normality, we want our Shapiro-Wilk statistic to be greater than 0.05 (not significant). In other words, we do not want our sample to differ significantly from the normal distribution
• Analyze > Compare means > Independent Samples T test


  Running our Independent t-test
• Reading our Output


  Descriptive Statistics: Check our group statistics table to screen for any data-input errors. Check Homogeneity of Variance: Check Levene’s Test for Equality of Variances. Check T-Test Statistics: Look at our t-test for Equality of Means
• Independent-samples t-test


  Conducted to compare the mental resilience of people who have, or have not, completed a Spartan Adventure race
• Checking Assumptions: Normality - Shapiro-Wilk statistic should be greater than 0.05 to meet the assumption of normality
• Paired Samples T test

  Running a Repeated Measures t-test
• Descriptive Statistics table helps screen for data-input errors and ensures correct number of participants
• 1. Test Statistics table shows t statistic, degrees of freedom, and significance value
• Paired-samples t-test


  Conducted to evaluate the effect of having a hangover on the number of hours spent watching daytime TV