2. Review
• Last class we learned:
– t-test – hypothesis testing where the population
________________- is unknown
– t-test for ______________- One sample vs
population
– Went through all the steps –
• Estimate new pop variance
• Estimate new DofM
• Use new type of sig. test (t-score)
3. t-test for Dependent Means
• t-test for Dependent Means – uses
___________________for each person and pop
_____________ is still not known; want to see
the ____________ in scores for
___________________________.
• This t-test uses a
__________________________design – each
participant is measured
________________, usually
______________________some
“____________________”
• The means are dependent on each
other, because they are both from the same
4. Calculation
• There are 2 differences between t-tests for
single samples and t-tests for dep. Means
– 1. __________________________– need to
subtract scores to get 1 score
• _________ score _____________ score = __________
(the difference) score
– 2. Assume new ________________– since we are
now using _____________________scores, and
the null hypothesis is that there is ______
difference between scores, we assume that the
_____________________is now __
5. Example
• Hypothesis: Statistics students who get help from
statistics tutors will have ___________ GPAs then
statistics students in general.
• 1. Begin with stating what your two populations are:
– _____________– Students who went to stats tutors’ GPAs
– _____________– All stats students’ GPAs
• 2. Next, state your hypotheses:
– __________________– there is_____difference between
students who went to stats tutors and all stats students.
– ___________________– there _____ a difference
between students who went to stats tutors and all stats
students.
• 3. Goal:
– What is the probability of getting certain results, if there
is no difference
6. • 4. Determine probability:
– _________
• 5. Determine your sample’s t-score (if sample
mean difference score is ___, and DofM SD is __)
t = M-u / Sm = _________ = ____ = __
• 6. Decide whether to reject or accept the null
hypothesis with a sample size of ____
7. df = ____
probability = ___
One or two tailed = ____
t-score = ___
Cut-off value = ____
_________
.05
1.71 2
8. • Can also have 2 scores from
__________people – if you want to examine
________________, _________, etc
• Called a ________________________
9. Assumptions
• Conditions/ __________________ for
hypothesis testing
• We have already talked about
________________ (ch 14)
– in order to make assumptions about our data
(the _________ is just like the pop;
_________________ to a population), we must
always assume the ________________ to be
______________, and make our ____________as
normal as possible
10. Assumptions cont.
• We will learn more assumptions later. . . BUT
• Assumptions are commonly
____________(data usually isn’t normal), so
we make corrections (such as ___________
estimates) to help with this problem
• T-test is very ____________ – good at
“_________________” for violated
assumptions, i.e. non-normal populations and
samples