week 11: t test for independent means Flashcards
The t test for independent means
used whenever two separate sample means are to be compared
when to use t tests for independent means examples
- where participants are randomly allocated to one of two groups or conditions (e.g. Control vs. Drug)
- situations where the variable that you are interested in studying cannot be manipulated (e.g., gender, marital status, handedness, personality dimensions) - need to find a way of comparing scores from people representing both levels of a variable (e.g. Male vs. Female)
what distribution is needed?
t distribution
Variance of Comparison Distribution
Step 1: Estimate Population Variance (SS1/df1 and SS2/df2
) then pooled together to make Ssquaredpooled
Step 2 Calculate the Variance of Each of the Distribution of Means
Step 3 Calculate the Variance and Standard Deviation of the Sampling Distribution of the Differences Between Means
Steps in the process of calculating independent groups t test
- State the hypotheses
- Determine Characteristics of Comparison Distribution
Mean of the distribution of the differences of means
Estimated population variance from both samples
The pooled estimate of the population variance
The variances of each distributions of means
The variance and standard deviation of the distribution of the differences between the means - Determine the cut-off score
- Calculate the t score
- Compare the position of your t score on the appropriate comparison t distribution based on df (if larger than .05 cut-off: reject H0; if smaller: retain H0)
The shape of the distribution of the differences between means
t distribution
the degrees of freedom are:
dftotal = df1+ df2 = (N1 − 1) + (N2 − 1) = N − 2
use dftotal for looking up critical values in the t table
Effect size in t tests
Effect size = strength of the relationship between IV and DV
Eta Squared η2
it gives us a measure of the amount of variability in the dependent variable that is related to the independent variable