Please enable JavaScript.
Coggle requires JavaScript to display documents.
T-tests (Background (3 varieties (One sample - t-test: compare data to…
T-tests
Background
The t-test is used to test hypotheses about means when the population variance is unknown (the usual case).
-
-
-
-
Effect size
-
Cohen's d : 0.2 indicates a small effect; 0.5 indicates a medium effect; 0.8 indicates a large effect
ONE SAMPLE/ Assumptions
- The sample is a random sample?
- The dependant variable is measured on an Interval or Ratio scale
- The Dependent variable is normally distributed?
Test the assumptions
- Values: how many times
- Normality is tested using a Kolmogorov-Smirnov test (50 or more values) or a Shapiro-Wilk test (fewer than 50 values).
- Convert to seconds
- We can't be 100% correct
- Example - Rowing: All time data to complete the 2000 m, compare to 420 seconds: We can test this using a one tailed one-sample t-test.
-