Class Notes (1,100,000)
CA (650,000)
SFU (10,000)
PSYC (1,000)
PSYC 210 (40)
Snjezana Huerta (20)
Lecture 17
PSYC 210 Lecture Notes  Lecture 17: Level Of Measurement, Standard Deviation, Test Statistic
by OC866704
School
Simon Fraser UniversityDepartment
PsychologyCourse Code
PSYC 210Professor
Snjezana HuertaLecture
17This preview shows half of the first page. to view the full 2 pages of the document.
Ñ° 210 Hypothesis Test examples
independent t test with homogeneity of variance
Snjezana.Huerta@sfu.ca
Does the level of academic achievement differ for firstborn children who are daughters and
firstborn children who are sons in Canadian families? (Assume that academic achievement is
normally distributed in both populations and that we are measuring it using an interval scale of
measurement.) Data were collected from 10 randomlyselected firstborn daughters and 10
randomlyselected firstborn sons. Use an alpha of .01.
Step 1: Specify the Parent Populations.
X
1
~
where 1 = firstborn daughters
X
2
~
where 2 = firstborn sons
Scale of measurement: ____________
Assumptions to be tested:
â€¢ Normality of the distribution of the X
1
scores
â€¢ Normality of the distribution of the X
2
scores
â€¢ Homogeneity of variance (i.e.,
22
2
2
1X
ÏƒÏƒÏƒ
==
)
Step 2: Specify hypotheses.
H
0
:
H
1
:
Step 3: Specify the test statistic and the null distribution.
21
)(
21
XX
s
cXX
t
âˆ’
âˆ’âˆ’
=
~
t
(
n
1
+
n
2

2
) n
1
= number of firstborn daughters, n
2
= the number of firstborn sons
(
)
==âˆ’Î•
âˆ’
21
21 XX
XX
Âµ
n =
ï£·
ï£·
ï£¸
ï£¶
ï£¬
ï£¬
ï£
ï£«+=
âˆ’
21
211
21
nn
ss p
XX
calculated after collect data
Step 4: Specify decision rule(s).
Î± = _____; ____tailed test;
df = _____________________________;
critical value(s): ________________________
*Collect data (using random sampling) & conduct thorough descriptive analyses
(see p. 2)
Step 5: Calculate test statistic.
ï£·
ï£·
ï£¸
ï£¶
ï£¬
ï£¬
ï£
ï£«+=
âˆ’
21
211
21
nn
ss p
XX
21
)(
21
XX
obs
s
cXX
t
âˆ’
âˆ’âˆ’
=
Step 6: Make a decision re: H
0
.
t
obs
t
crit
21
XX âˆ’
Âµ
You're Reading a Preview
Unlock to view full version