Sie sind auf Seite 1von 11

Cha

pter
8

Hypothesis Testing Z and t Tests

Group 1: Dara Permatasari


Diah Erawati
Endah Saputri
Hanifah Yonda Bestari
Nesya Sabrina Rahmadani
Pendidikan Biologi A 2014
Universitas Negeri Jakarta

8.1 Testing for the Difference Between Two


Proportions
Often you want to analyze differences
between two groups in the proportion of items
that are in aparticular category.
The sample statistics needed to analyze these
differences are the proportion of occurrences
in group 1 and the proportion occurrences in
group 2.
With a sufficient sample size in each group,
the sampling distribution of the differnce
between the two proportions approximately
follows a normal distiribution.

This table presents the joint responses of


whether a specific wafer good or bad and
whether that wafer had particles present it.
counts of particles found cross-classified by wafer condition
Wafer Condition
Good
Particles
Yes
Present
No

Total

Bad
14

50

36
400
320
80
334

Total

116

450


For

these data,
The proportion of good wafers without particle
is = 0.9581 and the proportion of bad wafers
without particle is = 0.6897
Because the number of good and bad wafers
that have no particles (14&36), the sampling
distribution for the difference between the two
proportions is approximately normally
distributed. The null and alternative hypotheses
are as follows:
H0 : 1 2 (no differences between the
proportions for the good and bad groups)
H1 : 1 2 ( a difference between the
proportions for the two groups)

You decide to use the critical value approach.


With a level of significance of 0.05, the lower
tail area is 0.025, and the upper tail rea is
0.25. using the cumulative normal distribution
table, the lower critical value of 0.025
corresponds to a Z value of -1.96, and a upper
critical value of 0.025 corresponds to Z value
of +1.96.

You
decide to use the p-value approach to
hypothesis testing. That the probability of
obtaining a Z value greater than 7.93 is
virtually = 0.05, rject the null hypothesis. You
report that there is evidence of a difference in
the proportion of good an bad wafers that
have no particles. Clearly, the good wafers are
much likely to have no oarticles than the bad
wafers.

The null and alternative hypotheses are as


follows:
H0 : 1 = 2 (no difference in the proportion
of heart attacks between the group that wa
given aspirin and the group that was given the
placebo)
H1 : 1 2 (a differenve in the proportion of
heart attacks between the two groups)

8.2 Testing for the Difference Between the


Means of Two Independent Groups

Statisticians distinguish between using two


indepedent groups, the meaning of groups as
previously defined in this chapter, and two
related groups, in which the observations are
matched according to a relevent characteristic
or repeated mesasurements of the same
items are taken. The most common test of
hypothesis used is pooled-variance t test.

Pooled-Variance
t Test

the hypthesis test for the difference between the


population means of two independent groups taht require
that the sample variances of each group be combined inti
one estimate of the variance common in the two groups.
For this test, the test statistic is based on the difference in
the sample means of the two groups, and the sampling
distribution for the difference in the two sample means
approximately follows the t distribution.
In a pooled variance t test, the null hypothesis no
difference in the means of two independent population is:
H0 : 2 (the two population means are equal)
And the alternative hypothesis is:
H1 : 2 (the two population means are not equal)

Pooled-Variance t test assumptions


in testing for the difference between the
means, you assume that the population from
which the two independt samples are drawn are
normally distributed with equal variances. For
situations in which the two populations have
equal variances, the pooled-variances t test
sensitive to moderate departures from this
assumption, provided that the sample sizes are
large. In such situations, the pooled-variance t
test can be used without serious effect on its
power.

8.3 The Paired t Test

The test for the pooled variance t test and


the difference between two population
proportions presented earlier in this chapter
were based on differences between two
indepemdent group . Often, data are obtained
from observations in which the groups are
related.
The null and alternative hypothesis can be
axpressed as follows: H0 : D = 0
and
H1 : D 0

Das könnte Ihnen auch gefallen