Chapter 14: Independent Group t-Test Flashcards
When do you go by correlated group design?
- Null- hypothesis population u and o are unknown
- differences between PAIRED scores are analyzed.
Experiment with two conditions?
- random sampling of subjects from population.
What are the characteristics of Independent groups design?
- ____ of subjects from population
- _____ assignment to each condition
- Each subject is tested how many times?
- What are analyzed?
- t-test analyzes what?
- Random sampling of subjects from population
- Random assignment to each condition
- No basis for pairing of scores
- Each subject is tested ONLY once
- Raw scores are analyzed
- t-test analyzes difference between sample means.
What are the alternative hypotheses for Directional and Nondirectional?
- U1 > U2 or U1<U2 (directional)
- U1 does not = U2 (nondirectional)
What are the null hypotheses for Directional and Nondirectional?
- U1 <= U2 ( directional )
- U1 = U2 (nondirectional)
Sample scores from condition 1 include:?
- U1= mean of pop that receives condition 1
- O1 = standard deviation of the pop that receives condition 1.
Sample scores from condition 2 include:?
- U2= mean of pop that receives condition 2
- O2= standard deviation of the pop that receives condition 2
What is the major assumption of independent groups design?(3)
- changing level of IV affects MEAN ONLY
- Does not affect standard deviation or variance
Therefore: O^2 1 = O^2 1 - If IV has a real effect the means will differ, not variance
L> IV assumed to equally affect each individual
Sampling distribution of the difference between sample means ( X1-X2):
- Draw all possible samples of __ and ___, from a population of __ and __, and standard deviation of __ and __.
- Calculate __ and __ for each sample
- Calculate __-__ for all possible pairings of samples of size __ and __.
- n1, n2, u1, u2, o1, o2
- Xbar 1, Xbar 2
- Xbar 1- Xbar 2
- n1 and n2
What is the formula for t-test of independent groups?
- Tobt = (X1-X2) - uX1- uX2/ square root of: S^2w( 1/n1 + 1/n2)
- ## S^2w = weighted estimate of o^2 (variance)
What is the computational equation for tobt???
- tobt= (X1-X2)/ square root: (SS1+SS2/n1+n2-2) times by (1/n1+1/n2)
What is the formula for df for independent groups?
- N-2
Conclusion for independent groups? (t)
- | tobt | >= | tcrit |, reject Ho
What is the formula for Tobt when n1=n2 ?
degrees of freedom ?
- Tobt= (X1-X2)/ square root of: SS1+SS2/ n(n-1)
- Df= N-2 (N= population NOT SAMPLE)
What is the assumption of independent group t-tests?
- test is robust
The sampling distribution is ___ distributed. If __=__ and >=.
- normally
- n1=n2
- n>= 30
Homogeneity of variance? huh?
- O1^2 = O2^2
L> Mann- Whitney U
What is the formula for effect size of independent groups?
- D= | X1- X2| / square root of: S2w
ala –> S2w= SS1+ SS2/ n1+n2-2
What increases power??
- Great effect of IV
- Larger sample size
- Decreased variability
Correlated t-tests is more powerful when correlation between paired scores is __.
- high
What is the equation for confidence intervals?
- 95%
- 99%
- u lower= (X1-X2)- Sx1-x2 t0.025
- u upper= (X1-X2) + Sx1 - X2 t 0.025
- u lower= (X1-X2)- Sx1-X2 t 0.005
- u upper= (X1-X2) + SX1- X2 T 0.005
SX1-X2=??
SQUARE ROOT OF: SS1+SS2/n1+n2-2)(1/n+1/n2)
When interpreting confidence intervals the probability value applies to the __ not the ___.
- interval
- population mean