TESTS OF

HYPOTHESIS
Objectives:
1. Illustrates the:
• Null hypothesis
• Alternative hypothesis
• Level of significance
• Rejection region
• Types of errors in hypothesis testing
2. Solves problems about tests concerning means
Important
Terms
Mean & sd
Normal Curve
 Parameter &
Statistic/Estimate
Introduction

• A hypothesis is a statement or a tentative

theory that may or may not be true,   but is
initially assumed to be true until new evidence
suggests otherwise. It may be   proposed from
a preliminary observation, a guess or based
from previous   experiences. 
Types  of  Hypothesis
• Null hypothesis – the null hypothesis is denoted by
Ho, it is the hypothesis of “no difference” and usually
formulated for the purpose of being rejected.
• 
• Alternative hypothesis – the alternative hypothesis
is   denoted by Ha or H1. This the hypothesis that
contradicts the hull hypothesis.   If the null
hypothesis is rejected,  the alternative is being
supported.  The alternative hypothesis is the
operational statement of the experimenter’s 
research hypothesis.
Types of Errors in Hypothesis Testing
• Two types of errors are usually committed in deciding
to accept or reject the null hypothesis
•   Type I error – usually committed if the Ho is rejected
when the Ho is true
•   Type II error – usually committed if the Ho is
accepted when the Ho is false

Status of Hypothesis Accept Ho Reject Ho

Ho is True Correct Decision Type I error
Ho is False Type II error Correct Decision
Example
• It was a stormy Wednesday night and
a student considered the null
hypothesis: There won’t be any
laboratory class in STAT class the next
morning. The next day, he came to
class all soaked up only to find the
classroom empty.
One-sided and Two-sided Test

•  In one-sided test the Ha specifies that the

unknown     population parameter is entirely
above or entirely below the specified     value
of the Ho. It is called one-tailed or a directional
test. While the     two- sided test, the Ha
specifies that the unknown population  
  parameter can lie on either side of the value
specified by Ho. It is     called   two-tailed or
nondirectional test.
Level of Significance
• The probability of committing type I error is
called the level of  significance or margin of
error of the test and it is  denoted by
α (alpha), and one minus the level of
significance is called confidence  level.
• The probability of committing type II error is
denoted by ß, and (1– ß) is called the power of
the test. This will indicate the ability
of the test statistics to determine correctly
that the Ho is false, hence it should be
rejected.
The Critical Region

• Critical region is the set of all

values of the test statistics
that would  cause to reject the null
hypothesis.
Basic Steps in Hypothesis Testing
• State the Ho and Ha, decide what data to collect
and under what conditions
• Specify  the level of significance  and the
sample size n
• Find the sampling distribution of the test
statistics under the assumption that Ho is true
• Establish the critical region for the test statistics
• Computation of the test statistics, for a sample
size n
• Decision.
General types of tests

• Parametric tests-tests that require

certain assumptions that must be met
• Nonparametric tests-test to examine
differences between groups on ordinal
data or skewed distributions of data
and test relationships between
categorical variables.
Comparing Sample and Population
The population variance is generally unknown and is estimated
by the variance of the random samples.  The sampling variability
of the sample variance may be affected if the sample size
is small (n less than or equal to 30).  Hence, the unavailability of
the population variance must take into consideration and to be
estimated by the sample variance.
          If the given observations X1, X2, . . ., Xn is a
random sample from a normal distribution, but the population
variance is unknown, then the test statistics

has a   t-distribution with  (n-1) degrees of freedom.

Sample Problem
A high school principal claims that the average
performance of his  graduating class in math was 83. 
To test this claim, 25 students were randomly
selected from the recent graduating class with
the  following results:    87,  85,  76,  83,  78,  90,  89, 
85,  82,  77,  79,76,  86, 83,  93, 88, 84,  76,  79,  85,  82, 
81,  85,  84,  80.   Test the principal's claim using 0.05
level of significance.
Answer
• 1.  Ho :     µ = 83
     Ha :     µ ≠ 83
•  2.   α  = 0.05,       n = 25
 3.   Test Statistics :     t-test
•  4.   Critical Region :   Reject the Ho if  |tc | > 2.064
•  5.   Computations:  _                                                       
                                                      X = 82.92    S = 4.6    n = 25
•                                                __
                                 (82.92– 83) V25
                        tc = ———————— = -0.09
                                          4.6
• 6.    Decision: Since tc < 2.064, therefore FtRHo and conclude 
that the claim of the principal is valid and the performance of
new graduating class in Math is not significantly different from
the previous graduates.
Let’s Practice!!!
Comparing Two Dependent Samples
• The difference between two populations when
dependent or correlated samples was used is almost
the same with the t-test with singles samples .
• The only difference is that in dependent samples we
will be dealing with the difference of two observed
values rather than the original values.

d-average observed difference

do-hypothesized average difference
Sd-standard deviation of observed
difference
N-number of paired values
Sample Problem:
The following data represent the weights (in lbs.) of 10
obese participants before and after undergoing  a
two-weeks  weight  reducing training program.
Participants 1 2 3 4 5 6 7 8 9 10
Before 275 250 235 212 215 263 289 258 215 249
After 250 215 218 198 200 245 260 250 202 230

Test the hypothesis that the mean weight loss of

obese participants     after the training program is 10
lbs. against the alternative that the loss  weight is
greater than 10 lbs.  Use 0.01 level of significance.
Answer:
• 1.   Ho :     d = 10 lbs
                  Ha :     d > 10 lbs.
  2.   α  = 0.01,       n = 10
 3.   Test Statistics :     t-test (dependent samples)
 4.   Critical Region :   Reject the Ho if  tc > 2.821
• 5.   Computations: 

Participants 1 2 3 4 5 6 7 8 9 10
Before 275 250 235 212 215 263 289 258 215 249
After 250 215 218 198 200 245 260 250 202 230
Difference 25 35 17 14 15 18 29 8 13 19

d = 19.30        S = 8.12 5      n = 10                          

                                    (19.30– 10.0) V10
                        tc = ——————————  = 3.62
                                              8.125
6. Decision: Since tc > 2.821, therefore reject the  Ho hypothesis and
conclude that the training program is effective in reducing more than 10 lbs.
Comparing Two Independent Samples
• Hypothesis testing about the differences of means
between two independent populations will involved
the t-distribution and the t-test statistics. When the
two samples are independent, and the two
independent populations are normally distributed and
the population variances are unknown but assumed
to be equal.  

Df= (n-2)
 Sample Problem:
A classroom teacher wishes to compare the
performance of students in statistics using two
methods of teaching.  Two independent samples of
sizes 15 and 10 were randomly selected.   The
following data have been obtained: 
Method A 85 83 76 78 82 81 86 75 77 83
Method B 85 78 83 90 89 87 84 83 79 82 76 84 89 78 87

Is there a significant difference between the

performance of students  in the two methods of
teaching statistics? Use 0.05 level
of significance.
Answer:
• 1.   Ho :  μ1 = μ2
           Ha :   μ1 ≠ μ2
•  2.   α  = 0.05,       n1 = 10       n2 = 15

 3.   Test Statistics :     t-test (independent samples)

•  4.   Critical Region :   Reject the Ho if  I tcI > 2.069  
•  5.   Computations: 
• x1 = 80.60    S1 = 3.86                x2 = 83.60    S2 = 4.29
•                                      tc =  -1.753                         
• 6.  Decision:  Since I tcI < 2.069, therefore accept the 
Ho hypothesis and conclude that the two methods
have the same performance. 
Thank you for listening!!!