Académique Documents
Professionnel Documents
Culture Documents
(CORRELATED)
DESIGN
if
members
of
one
sample
can
be
used
to
determine
the
members
of
the
other
sample
when
there
is
a
matching
between
the
subjects
in
one
sample
and
the
subjects
in
the
other
samples
(matched-pairs
designs
or
correlated)
e.g.
husbands
and
wives;
twins
or
paired
with
age,
IQ
we
measure
every
participant
more
than
once
(repeated
measures
design)
e.g.
pretest-posttest
look
for
changes
that
occur
within
participants
as
the
same
people
are
tested
in
different
conditions
(within-subjects
design)
e.g.
causal
effects
(before-after)
benefit
of
controlling
for
some
of
the
extraneous
factors
because
participants
acts
as
their
own
control
groups
are
already
matched
making
them
always
equal
requires
fewer
participants
because
of
equality,
they
are
better
able
to
detect
differences
Dependent
Designs
are
prone
to
Order
Effects:
Type of Effect Practice Effect Definition Improvements in task performance due to learning or warming up. (Novice golfer hits balls more and more accurately over time of practice.) Decrease in task performance due to becoming tired or bored. The influence of a treatment or condition extends long enough to affect behavior in a subsequent condition. Can improve or impair performance. Control/Minimize Give subjects practice on the task before introducing the research conditions. Make the task as brief and interesting as possible. Allow enough time after each condition for its influence to wear off.
! Statistical analysis does not examine the original scores of participants instead, analyzes difference between scores of the two conditions hypothesis are made about the difference scores Ho = the mean difference is equal to zero Two-tailed: Ho : xd= 0 One-tailed: Ha : xd0 or xd0 Assumptions data chosen used for comparing the mean must be identical or matched subjects over a period of time or in a different circumstances. population from which the observed data is taken follows a normal distribution. The standard deviation of both the group must be approximately equal.
When to use paired t-test: When there are two samples and a within-groups design When sampling distribution is normally distributed When dependent variable is on interval or ratio scale FORMULAS:
STEPS: STEP 1; Assume the Ho and that the mean difference is zero STEP 2: Find the differences between the pair observation. STEP 3:Find the mean difference d. STEP 4: Find the Standard Deviation. STEP 5: Find the t-statistic value. STEP 6: Use the table of t-distribution and compare df. STEP 7: Make decision about the Ho. (Reject/Failure to Reject). SAMPLE PROBLEM Does attending a mixed race day camp affect childrens racial attitude? Six children attending a day camp ware given a test to measure racial prejudice (higher scores = more prejudice) when they first arrived at camp (Condition A: Before Day Camp). The same six children were given the same test later when they left camp (Condition B: After Day Camp).
Attending mixed race camps greatly affects prejudice levels of children. Sample problems A researcher is interested in changes over time in happiness after a break up. He asks 6 people to complete a Happiness scale the first day after a break up and then complete the same scale again 20 days later. Higher numbers represent more happiness.
The following sets of scores were made by 10 individuals in lab experiment on perception. Is there a significant difference between the means of the distributions. Alpha level is .05.