Ch. 13 - The Comparison of Two Conditions
Tuesday, February 26, 20132:17 PM
What do signal-to-noiserationshaveto do with t tests?
• Three most popular statistical tests: t test, F test, and chi-square test
• The hypothesis determines which statistical test is used
○ t test-- used for comparing the means of two groups
• A signal-to-noise ratio is set up
○ Signal is represented by the difference between the two means
○ Noise is represented by the variability of the scores within the samples
• The larger this ratio, them more likely it is that the null hypothesis will be rejected
• The independent-sample t test tests statistical significance by examining the difference between two
independent means (the signal) against the background of the within-group variability (the noise)
• The larger the difference between the means (i.e., the greater the signal), and/or the smaller the within-group
variability for a given size study (i.e., the less the noise), the greater will be the value of t (larger t value is
How do I computean independent-sample t test?
• t test for independent samples (i.e., not repeated-measures designs)
• Where (unbiased estimator of the population variance)
What cana tableof p valuesfor t teachme?
• t distribution-- a family of curves, where there is a different curve (each one resembling the standard normal
distribution) for every possible value of what are called the degrees of freedom (df)-- the number of values free
○ For an independent-samples t test with two groups,
• In order for a t test to be accurate, homogeneity of variance (the idea that the population variance of the groups
being compared is assumed to be equal) must be met
○ Raw scores are often tr