Understanding and Applying Statistical Tests

Introduction
Statistical tests play a crucial role in research and data analysis. Choosing the
appropriate test is essential for drawing valid conclusions. This module aims to
provide an overview of common statistical tests and guidelines for their proper
application.

Lesson 1: Tests for Significant Relationships

Overview
Tests for significant relationships assess how variables are related to each other.

Pearson’s Correlation Coefficient

Description: Measures the strength and direction of the linear relationship
between two variables.
Type: Parametric.
Key points:
• Assumes data is normally distributed.
• Calculated by finding the slope of the line of best fit.
• The slope indicates the strength and direction (positive or negative) of the
relationship.
• Ranges from -1 to 1. (0 = no linear relationship, 1 = perfect positive
linear relationship, -1 = perfect negative linear relationship)
Formula: P
(x − x̄)(y − ȳ)
r = pP pP
(x − x̄)2 (y − ȳ)2
where:
• r is the Pearson’s correlation coefficient
• x is the value of the first variable

1
 • y is the value of the second variable
• x̄ is the mean of the first variable
• ȳ is the mean of the second variable

•
P
is the sum of the values
Steps:
1. Collect data.

2. Plot data.
3. Calculate the coefficient.
4. Interpret.

Spearman’s Rank Correlation Coefficient

Description: Measures the strength and direction of monotonic relationships.
Key points:
• Monotonic relationship: variables move in the same direction.

• Calculated using ranks of data.

• Ranges from -1 to 1.
Formula:
6 d2
P
ρ=1−
n(n2 − 1)
where:
• ρ is the Spearman’s rank correlation coefficient
• d is the difference between the ranks

• n is the number of observations

Steps:
1. Collect data.
2. Rank data.

3. Calculate the coefficient.

4. Interpret.

2
Chi-Square Test of Independence
Description: Tests if two categorical variables are independent.
Key points:

• Categorical variable: variable divided into categories.

• Uses the chi-square statistic.
Formula:
X (O − E)2
χ2 =
E
where:
• χ2 is the chi-square statistic

• O is the observed frequency

• E is the expected frequency
Steps:

1. Collect data.
2. Create a contingency table.
3. Calculate the statistic.
4. Compare to critical value.

5. Interpret.

Lesson 2: Tests for Significant Difference

Overview
Tests for significant difference compare groups or samples to see if there are
significant differences.

T-test
Description: Compares the means of two groups.
Type: Parametric.
Key points:
• Assumes data is normally distributed.
• Uses the t-statistic.

3
 Formula:
x̄1 − x̄2
t= q
s̄ n11 + n12
where:
• t is the t-statistic
• x̄1 is the mean of the first group
• x̄2 is the mean of the second group
• s̄ is the pooled standard deviation
• n1 is the number of observations in the first group
• n2 is the number of observations in the second group
Steps:
1. Collect data.
2. Calculate mean and standard deviation.
3. Calculate the t-statistic.
4. Compare to critical value.
5. Interpret.

ANOVA
Description: Compares the means of three or more groups.
Type: Parametric.
Key points:
• Assumes data is normally distributed.
• Uses the F-statistic.
Formula:
M SB
F =
M SW
where:
• F is the F-statistic
• M SB is the mean square between groups
• M SW is the mean square within groups (or error)
Steps:
1. Collect data.

4
 2. Calculate mean and standard deviation for each group.
3. Calculate Sum of Squares Between (SSB), Sum of Squares Within (SSW),
and Total Sum of Squares (SST).

4. Calculate degrees of freedom: df between = number of groups - 1, df within

= total number of observations - number of groups.
5. Calculate mean squares: MSB = SSB / df between, MSW = SSW /
df within.
6. Calculate the F-statistic.

7. Compare to critical value.

8. Interpret.

Chi-Square Test of Homogeneity

Description: Tests if two or more categorical variables have the same distri-
bution.
Key points:
• Uses the chi-square statistic.

Steps:
1. Collect data.
2. Create a contingency table.

3. Calculate the statistic.

4. Compare to critical value.
5. Interpret.

Lesson 3: Non-Parametric Tests

Overview
Non-parametric tests are used when data does not meet the assumptions of
parametric tests (e.g., normal distribution).

5
Mann-Whitney U Test
Description: Compares means of two independent groups.
Type: Non-parametric.
Key points:
• Does not assume normal distribution.
• Uses ranks.
Formula:
U = n1 (n1 + 1)/2 − R1
where:

• U is the Mann-Whitney U statistic

• n1 is the number of observations in the first group
• R1 is the sum of the ranks in the first group
Steps:

1. Collect data.
2. Rank data.
3. Calculate the statistic.

4. Compare to critical value.

5. Interpret.

Wilcoxon Signed-Rank Test

Description: Compares two paired samples.
Type: Non-parametric.
Key points:
• Substitute for paired t-test.

• Does not assume normal distribution.

Formula:
W = n(n + 1)/4 − T
where:
• W is the Wilcoxon signed-rank statistic
• n is the number of observations

• T is the sum of the positive ranks

6
 Steps:
1. Collect data.
2. Rank data.

3. Calculate the statistic.

4. Compare to critical value.
5. Interpret.

Kruskal-Wallis 1-Way ANOVA

Description: Compares means of three or more independent groups.
Type: Non-parametric.
Key points:

• Substitute for one-way ANOVA.

• Does not assume normal distribution.
Formula:
12 k(k + 1)
H= (R(1) + R(2) + ... + R(k)) −
N 4
where:
• H is the Kruskal-Wallis statistic

• N is the total number of observations

• R(i) is the sum of the ranks for group i
• k is the number of groups

Steps:
1. Rank all data.
2. Assign ranks to groups.
3. Calculate the H-statistic.

4. Compare to critical value.

5. Interpret.

7
Friedman 2-Way ANOVA
Description: Compares means of three or more related groups.
Type: Non-parametric.
Key points:
• Substitute for repeated measures ANOVA.
• Does not assume normal distribution.
Formula:
12 k(k + 1)(2k + 1)
X2 = (R(1) + R(2) + ... + R(k)) −
N 6
where:
• X 2 is the chi-square statistic
• N is the total number of observations
• R(i) is the sum of the ranks for group i
• k is the number of groups
Steps:
1. Rank data within each group.
2. Calculate sum of ranks.
3. Calculate chi-square statistic.
4. Compare to critical value.
5. Interpret.

Lesson 4: Choosing the Appropriate Statistical

Test
Overview
Selecting the correct test depends on the research question and data character-
istics.

Key Considerations
• Type of data (nominal, ordinal, interval/ratio)
• Number of groups/samples
• Looking for differences or relationships
• Assumptions of parametric tests

8
Table of Appropriate Tests
Looking for Type of Data Comparing
Difference of Parametric 2 groups/samples: t-test
groups/samples for Independent samples,
t-test for paired samples
3 or more groups:
ANOVA
Non-Parametric 2 groups/samples:
Mann Whitney U-Test,
Wilcoxon Signed Rank
Test
3 or more groups:
Kruskal-Wallis 1-way
ANOVA, Friedman 2-way
ANOVA
Relationship of Parametric (Inter- Pearson r, Linear Regres-
variables val/Ratio) sion
Non-Parametric Chi-square
(Nominal)
Non-Parametric Spearman Rank
(Ordinal)

Conclusion
Understanding and correctly applying statistical tests is crucial for drawing ac-
curate and reliable conclusions from data. This module has provided an overview
of common tests and guidelines for their appropriate use.