Hypothesis Testing

Hypothesis
A hypothesis is a statement which
expresses an idea or concept which
needs to be proven true or false.
Types of Hypotheses
Null Hypothesis
Alternative Hypothesis
Null Hypothesis, H0
The null hypothesis is denoted by H0
It expresses the idea of non-significant difference.
It is expressed as an equality.
It states that a given distribution equals the norm
or generally accepted standard.
It states that two distributions are equal when
compared with each other.
Alternative Hypothesis, Ha or H1
The alternative hypothesis is denoted by Ha or H1
It expresses the idea of significant difference.
It is expressed as an inequality using the ≠, >, or <
It states that a given distribution is not equal to, greater
than or less than the norm or generally accepted
standard.
It states that a distribution is not equal to, greater than
or less than another distribution when compared with
each other.
Types of Errors
Type I
• Type I error happens when the null hypothesis
is rejected when it is true.
Type II
• Type II error happens when the null hypothesis
is accepted when it is false.
Level of Significance
The probability of committing a type I error is
called the level of significance.
The level of significance for a type I error is
denoted by alpha, α
The probability of committing a type II error is
denoted by beta, β
Type of Test
The type of test to use is based on the
alternative hypothesis. The types of tests are:
1. Two-tailed Test
2. One-tailed Test
Two-Tailed Test
The two- tailed test also known as the non-
directional type of test is used when the
researcher is only interested in knowing
whether there is a difference between the
groups being compared.
It is indicated by the use of the relational
operator, ≠ , in the alternative hypothesis, Ha .
Two-Tailed Test
One-Tailed Test
The one- tailed test also known as the
directional type of test is used when the
researcher is interested in knowing whether
one group is better or worse than another
group.
It is indicated by the use of the relational
operators, < or > , in the alternative
hypothesis, Ha .
One-Tailed Test with a Negative
Critical Value
The one- tailed test with a negative critical
value is used when the researcher is
interested in knowing whether a first group
is worse or less than a second group.
It is indicated by the use of the relational
operators, < , in the alternative hypothesis,
Ha .
One-Tailed Test with a
Negative Critical Value
One-Tailed Test with a Positive
Critical Value
The one- tailed test with a positive critical
value is used when the researcher is
interested in knowing whether a first group
is better or greater than a second group.
It is indicated by the use of the relational
operators, > , in the alternative hypothesis,
Ha .
One-Tailed Test with a
Positive Critical Value
Type of Statistical Test
1. z - Test
2. t - Test
3. F – Test also known as Analysis of Variance
or ANOVA
z - Test
Used to compare a group with the existing
norm or standard.
Used to compare two groups.
Used when the standard deviation for
population, σ , is given or known.
May be used when σ is unknown or not given
but the sample size, n, is sufficiently large.
n ≥ 30
t - Test
Used to compare a small group with the
existing norm or standard.
Used to compare two small groups.
Used when the standard deviation for
population, σ , is unknown or not given
and the sample size, n, is small, n < 30, or
when the sample sizes, n1 and/or n2 are
small, n1 , n2 < 30
F – Test (ANOVA)
Used to compare three or more groups.
If there is no difference between the groups
being compared, make your conclusion.
If there is a difference between the groups
being compared then additional tests need to
be performed to determine where the
differences lie between the groups.
z – Test Formulas
Comparison Between Sample Mean and Existing Norm/Standard
𝑥−𝑢
𝑧= 𝜎
𝑛
Where
z = z – test value
x = sample mean
μ = population mean
σ = population standard deviation
n = sample size (number of cases/observations)
Comparison Between Sample Proportion and Population
Proportion (Existing Norm/Standard)
𝑃−𝑝
Where 𝑧=
𝑝𝑞
z = z – test value 𝑛
P = sample proportion
p = population proportion
q=1-p
n = sample size (number of cases/observations)
 Comparison between the Means of Two Groups
𝑥1 − 𝑥2
𝑧=
Where 1 1
𝜎 +
z = z – test value 𝑛1 𝑛2
𝑥1 = mean of 1st group
𝑥2 = mean of 2nd group
σ = population standard deviation
n1 = sample size of 1st group
nd
n2 = sample size of 2 group
 Comparison between Two Population Proportions
𝑝1 − 𝑝2
𝑧=
Where 𝑝1 𝑞1 𝑝2 𝑞2
+
z = z – test value 𝑛1 𝑛2
𝑝1 = proportion of 1st group
𝑝2 = proportion of 2nd group
𝑞1 = 1 - 𝑝1
𝑞2 = 1 - 𝑝2
n1 = sample size of 1st group
n2 = sample size of 2nd group
t – Test Formulas
Comparison Between Sample Mean and Existing Norm/Standard
𝑥−𝑢
t= 𝑠
𝑛
Where
t = t – test value df = n - 1
x = sample mean
μ = population mean
s = sample standard deviation
n = sample size (number of cases/observations)
 Comparison between the Means of Two Groups, equal σ2

Where
𝑥1 − 𝑥2
t=
t = t – test value 1 1
𝑠 𝑝 +
𝑥1 = mean of 1st group 𝑛1 𝑛2
𝑥2 = mean of 2nd group
𝑛1 − 1 𝑠12 + 𝑛2 − 1 𝑠22
𝑠𝑝 = pooled standard deviation 𝑠𝑃 =
𝑛1 + 𝑛2 − 2
𝑠1 = standard deviation of 1st group
𝑠2 = standard deviation of 2nd group ⅆ𝑓 = 𝑛1 + 𝑛2 − 2
n1 = sample size of 1st group
n2 = sample size of 2nd group
 Comparison between the Means of Two Groups, unequal σ2
𝑥1 − 𝑥2
𝑡=
Where 2 2
𝑠1 𝑠2
t = t – test value +
𝑛1 𝑛2
𝑥1 = mean of 1st group
𝑥2 = mean of 2nd group Satterthwaite Approximation Formula

𝑠1 = standard deviation of 1st group 2

𝑠2 = standard deviation of 2nd group 𝑛1 + 𝑛2

𝑥𝑑 − 𝑥𝑑 2
𝑠𝑑 =
𝑛−1