You are on page 1of 5

Sigmath Reviewer

STATISTICS QUARTER 4
I. Chi-Square Test

- The chi-square test is a statistical method used to examine the relationship between two
categorical variables. It is commonly used in research studies to test whether there is a significant
association between the variables of interest. The test can be applied to various types of data,
including survey data, medical data, and marketing data.

- The chi-square test involves calculating the difference between the observed and expected
frequencies of the variables of interest. The observed frequency refers to the actual number of
times a value occurs for each variable, while the expected frequency refers to the number of
times a value is expected to occur based on the null hypothesis.

- There are two types of chi-square tests: the goodness-of-fit test and the test for independence.
The goodness-of-fit test is used to determine whether the observed data fits the expected
distribution, while the test for independence is used to determine whether two categorical
variables are independent or dependent.

Formula:

Σ((O-E)²/ E)

where:
- χ² is the chi-square test statistic
- O is the observed frequency
- E is the expected frequency

STEPS:
- State the null hypothesis (H0) and the alternative hypothesis (Ha)
- Look up the critical value, level of significance, and degrees of freedom.
- Statistical Test
- Test Statistic
Σ((O-E)²/ E)
- Decision and Interpretation
Sigmath Reviewer

II. Sign Test

- The sign test is used to determine if there is a significant difference between the mean
characteristics of two populations.

- Responses on each pair of A and B are compared.

- The number of times A exceeded B is used as the test statistic. It is denoted as letter X.

- This is called the sign test because X is the number of positive (or negative) signs associated
with the difference between the pairs in population A and B.
Sigmath Reviewer
- In such cases, the implied null hypothesis is that the two population distributions are identical or
there is no significant difference between the two population distributions A and B.

- For any given pair, the probability that A exceeds B is p=0.05 when the null hypothesis is true.

STEPS:
- State the null and alternative hypotheses.
- Determine the critical values.
- Test Statistic
Examine each pair (A and B) of observations. If A>B, assign a plus (+) sign, if A<B, assign a
minus (-) sign, if A=B, the pair must be discarded.
Count the number of pairs remaining and denote it by n.
Count the number of times the less frequent sign occurs and denote this by r.
- Decision
- Interpretation

III. Wilcoxon Signed Rank Test

- The sign test is very simple to use; however, it is not the most efficient distribution-free test
available.

- Wilcoxon Signed Rank Test is a better test because it takes into account the magnitude of the
observed differences between paired values in the two samples.

- The original signs of the differences are given to their respective ranks, hence, it is also called
signed rank test.

STEPS:
- State the null and alternative hypotheses.
- Determine the level of significance and critical values.
- Identify the statistical test to be used.
- Test statistic
Obtain the differences between paired values in two samples/treatments.
Rank the absolute values of these differences, giving the smallest difference rank 1 and
assigning average ranks to tied differences.
Get the sum of the positive ranks and the sum of the negative ranks.
Find the smaller of the absolute values of the two sums of the ranks and denote it by T.
- Decision and Interpretation
Sigmath Reviewer
IV. KRUSKAL WALLIS H TEST

- If there are more than two populations to be compared, the Kruskal Wallis H Test can be used.
This is very useful for completely randomized design (CRD) using independent random samples.

STEPS:
- State the null and alternative hypothesis.
- Level of Significance, df=k-1, critical/tabular value (𝑋^2)
- Statistical Test
- Test Statistic
𝐻=12/(𝑛(𝑛+1))(∑〖𝑇2/𝑛)〗−3(𝑛+1)
- Decision and Interpretation

ASSUMPTIONS:
- All sample sizes are greater than or equal to 5
- Tied ranks will assume the average of the ranks that they would have occupied if they had not
been tied.

V. FRIEDMAN TEST

- The Friedman Test is used when comparing more than two populations or treatments randomly
assigned within blocks.

- This is the counterpart of the F test or ANOVA used in parametric statistics.


Sigmath Reviewer
- Unlike ANOVA which is used with the assumption that the observations taken from each of the
populations are normally distributed, the Friedman test is used when no distributional
assumptions are necessary.

STEPS:
- State null and alternative hypotheses.
- LOS, df=t-1, Use the Chi-Square table for the critical/tabular value.
- Friedman F-test.
- Test Statistic
Let t represent the treatments, and b represent the blocks. Arrange the recorded
observations for each treatment-block combination in a two-way table in which the
treatments are placed in columns and the blocks in rows.
Rank data within the blocks, giving the lowest value rank 1 and tied values the average of
the two ranks.
Get the sum of the ranks in each column or per treatment.
𝐹c =12/𝑏𝑡(𝑡+1) (∑𝑇2)−3𝑏(𝑡+1)
- Decision and Interpretation

You might also like