Scribd
Upload
368 views11 pages

T-Test or Z-Test Decision

1) The document discusses t-tests and z-tests for comparing means and proportions. It provides the formulas and assumptions for these statistical tests. 2) Examples are given to demonstrate hypothesis testing for differences in means between two populations or groups using a z-test when samples are large, and a t-test when samples are small. 3) Guidance is provided on when to use a z-test versus a t-test based on whether the population standard deviation is known and the sample size.

Uploaded by

kami_r202
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
368 views11 pages

T-Test or Z-Test Decision

1) The document discusses t-tests and z-tests for comparing means and proportions. It provides the formulas and assumptions for these statistical tests. 2) Examples are given to demonstrate hypothesis testing for differences in means between two populations or groups using a z-test when samples are large, and a t-test when samples are small. 3) Guidance is provided on when to use a z-test versus a t-test based on whether the population standard deviation is known and the sample size.

Uploaded by

kami_r202
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

t-Test for Mean (sect 8.

4)

When we have small sample size (n < 30), and


population standard deviation (σ) is unknown
to us. We will use:

Test Statistic formula:


X̄ − µ
t= √
s/ n

Assumptions for t-test to be valid:

1. random sample drawn from population


2. the population is (approximately) normaly
distributed.

1
H0: µ = number (null hypothesis)
H1: Three variations (alternative hypothesis)

1. µ > number (upper-tailed)

2. µ < number (lower-tailed)

3. µ 6= number (two-tailed)

Note:
1. t is the test statistic from the formula
2. tα and tα/2 are values from t-distribution
table with n-1 degrees of freedom

2
Ex: Eercise 8.4, 8

Urban Canadians spent on average of 62.1 min-


utes commuting to and from work. A chamber
of commerce executive claims that the average
commute time in her city is less than the Cana-
dian’s average. She randomly selects 25 com-
muters and finds the average commute time is
53.9 minutes with a sample standard deviation
of 12.9 minutes. At α = 0.10, test the execu-
tive’s claim.

3
Ex: Exercise 8.4, 12

A university reports that the mean salary of


student parents is $91,600. The president from
another university randomly surveys 28 families
and finds that their average income is $88,500
with sample standard deviation $10,000. Can
the president conclude that there is a differ-
ence? At α = 0.05, is he correct?

4
When to use Z test or t test?

1. σ known or n ≥ 30, use


X̄ − µ
z= √
σ/ n

2. σ unknown and n < 30, use


X̄ − µ
t= √
s/ n

5
z-test for proportions (sect 8.5)

Test Statistic:
p̂ − p
Z=q
p(1 − p)/n

where

p̂ = X
n
p = population proportion
n = sample size

6
H0: p = number (null hypothesis)
H1: Three variations (alternative hypothesis)

1. p > number (upper-tailed)

2. p < number (lower-tailed)

3. p 6= number (two-tailed)

Note:
1. Z is the test statistic from the formula
2. Zα and Zα/2 are values from Z score table

7
Ex: Exercise 8.5, 18

A survey of earned doctorates in Canadian uni-


versities claims that 64% of them received teach-
ing assistantships. A university dean feels that
this figure is low for her province. A random
sample of 50 PhD students finds that 27 have
assistantships. At α = 0.05, is the dean cor-
rect?

8
Testing the Difference Between Two Means:
Large Samples (Section 9.2)

There are situations where we are interested


in comparing two groups. How could we make
conclusions that one group is different from
another group?

Ex: Exercise9.2, 10
We are interest if there is a difference between
Calgary and Edmonton regarding house price.
We randomly sample 35 houses in Calgary, and
find out that mean house price is $358,103
with sample standard deviation $21,471. We
randomly sample 40 houses in Edmonton, and
find out that mean house price is $303,304
with sample standard deviation $14,636.

Can we conclude that Calgary has a higher av-


erage house price than Edmonton, simply be-
cause $358, 103 > $303, 304?
9
Z test for comparing two means
(X̄1 − X̄2) − (µ1 − µ2)
Z= r
σ12 σ22
n1 + n2

Note:
1. X̄1 and X̄2 are means from two samples
2. µ1 and µ2 are population means of two
groups
3. n1 and n2 are sizes of two samples

Assumptions for the Z test to be valid:


1.Two samples must be independent of each
other.
2.Large sample (n1 ≥ 30, and n2 ≥ 30). If
σ1, σ2 not known, use s1, s2.

10
Ex: cont
Is there enough evidence to reject the claim
that the vaerage cost of a home in both loca-
tions is the same? Use α = 0.01

11

You might also like