Two-Way Table: A table that displays the frequency of two categorical variables. Marginal Frequency: The totals for each category along the margins of the table.
- Fill in known values
- Use row/column totals to find missing values
- Verify that grand total equals total sample size
- Calculate marginal frequencies
| Math | Science | Total | |
|---|---|---|---|
| Freshman | 45 | 55 | 100 |
| Sophomore | 60 | 40 | 100 |
| Total | 105 | 95 | 200 |
Row Marginals: Freshman total = 100, Sophomore total = 100
Column Marginals: Math total = 105, Science total = 95
Grand Total: 200 students surveyed
Row totals: 100 + 100 = 200 ✓
Column totals: 105 + 95 = 200 ✓
This table shows the distribution of preferences across grade levels. 105 students prefer Math, 95 prefer Science. Equal numbers of Freshmen and Sophomores were surveyed.
The completed two-way table shows marginal frequencies: 105 students prefer Math, 95 prefer Science; 100 are Freshmen, 100 are Sophomores. The grand total confirms 200 students surveyed.
• Table Completion: Use known totals to find missing values
• Marginal Frequencies: Row and column totals along table margins
• Consistency Check: Row totals and column totals must equal grand total
Joint Relative Frequency: Individual cell frequency divided by grand total. Marginal Relative Frequency: Row/column total divided by grand total.
Freshman-Math: 45/200 = 0.225 = 22.5%
Freshman-Science: 55/200 = 0.275 = 27.5%
Sophomore-Math: 60/200 = 0.30 = 30%
Sophomore-Science: 40/200 = 0.20 = 20%
Math total: 105/200 = 0.525 = 52.5%
Science total: 95/200 = 0.475 = 47.5%
Freshman total: 100/200 = 0.50 = 50%
Sophomore total: 100/200 = 0.50 = 50%
| Math | Science | Total | |
|---|---|---|---|
| Freshman | 0.225 (22.5%) | 0.275 (27.5%) | 0.50 (50%) |
| Sophomore | 0.30 (30%) | 0.20 (20%) | 0.50 (50%) |
| Total | 0.525 (52.5%) | 0.475 (47.5%) | 1.0 (100%) |
All joint relative frequencies should sum to 1 (or 100%): 0.225 + 0.275 + 0.30 + 0.20 = 1.0 ✓
The relative frequency table shows that 22.5% of students are Freshmen who prefer Math, 30% are Sophomores who prefer Math, etc. All relative frequencies sum to 100%.
• Relative Frequency: Individual frequency ÷ Grand total
• Joint Relative Frequency: Each cell value ÷ Grand total
• Sum Check: All joint relative frequencies sum to 1 (or 100%)
Conditional Relative Frequency: The ratio of a joint frequency to its marginal frequency. It shows the proportion of one category within another category.
Freshman given Math: 45/105 = 0.429 = 42.9%
Sophomore given Math: 60/105 = 0.571 = 57.1%
Math given Freshman: 45/100 = 0.45 = 45%
Science given Freshman: 55/100 = 0.55 = 55%
Given Subject (Column Conditional):
| Given Math | Freshman | Sophomore |
|---|---|---|
| Frequency | 45/105 = 0.429 | 60/105 = 0.571 |
Given Grade Level (Row Conditional):
| Given Freshman | Math | Science |
|---|---|---|
| Frequency | 45/100 = 0.45 | 55/100 = 0.55 |
What percentage of Math students are Freshmen? 42.9% (45/105)
What percentage of Freshmen prefer Math? 45% (45/100)
42.9% of Math students are Freshmen, and 45% of Freshmen prefer Math. These are different conditional probabilities showing different perspectives on the same data.
• Conditional Frequency: Joint frequency ÷ Relevant marginal frequency
• Direction Matters: "Given Math" vs "Given Grade" yield different results
• Sum to One: Conditional frequencies in each group sum to 1 (or 100%)
Two-Way Table: A table that displays the frequency distribution of two categorical variables.
Joint Frequency: The count in each individual cell of the table.
Marginal Frequency: The totals for each row and column, found in the margins of the table.
Joint Relative Frequency: The proportion of the grand total represented by each individual cell.
Marginal Relative Frequency: The proportion of the grand total represented by each row/column total.
Conditional Relative Frequency: The proportion of one category within a specific subset of another category.
- Organize Data: Place categorical data in rows and columns
- Complete Table: Fill in all cells including row/column totals
- Calculate Frequencies: Find joint, marginal, and conditional frequencies
- Interpret Results: Analyze patterns and relationships between variables
- Draw Conclusions: Use frequencies to make data-driven statements
• Table Completeness: Every cell, row total, column total, and grand total must be filled
• Consistency Check: Sum of row totals = Sum of column totals = Grand total
• Relative Frequency: Joint Relative Frequency = Cell Value ÷ Grand Total
• Conditional Frequency: Conditional Relative Frequency = Joint Frequency ÷ Relevant Marginal Frequency
• Probability Interpretation: Relative frequencies can be interpreted as probabilities
Association: A relationship between two variables where knowing the value of one variable provides information about the value of the other variable.
| Passed | Failed | Total | |
|---|---|---|---|
| Studied | 80 | 10 | 90 |
| Did Not Study | 30 | 30 | 60 |
| Total | 110 | 40 | 150 |
Pass rate for those who studied: 80/90 = 0.889 = 88.9%
Pass rate for those who didn't study: 30/60 = 0.50 = 50%
Students who studied have an 88.9% pass rate, while those who didn't study have only a 50% pass rate.
Since 88.9% ≠ 50%, there appears to be a strong positive association between studying and passing the test.
Students who studied were almost twice as likely to pass the test compared to those who didn't study, indicating a strong positive association.
There is a strong positive association between studying and passing the test. Students who studied had an 88.9% pass rate compared to only 50% for those who didn't study. The conditional relative frequencies show a significant difference, indicating association.
• Association Detection: Compare conditional relative frequencies between groups
• Strong Association: Large differences in conditional frequencies
• No Association: Conditional frequencies approximately equal
Statistical Association: A relationship between two variables where the conditional distributions differ significantly from the marginal distributions.
| Satisfied | Dissatisfied | Total | |
|---|---|---|---|
| Received Offer | 200 | 50 | 250 |
| Did Not Receive Offer | 120 | 30 | 150 |
| Total | 320 | 80 | 400 |
Received offer and satisfied: 200/400 = 0.50 = 50%
Received offer and dissatisfied: 50/400 = 0.125 = 12.5%
No offer and satisfied: 120/400 = 0.30 = 30%
No offer and dissatisfied: 30/400 = 0.075 = 7.5%
Satisfaction rate for those who received offers: 200/250 = 0.80 = 80%
Satisfaction rate for those who didn't receive offers: 120/150 = 0.80 = 80%
Both groups have the same satisfaction rate (80%), suggesting no association between receiving an offer and customer satisfaction.
Overall satisfaction rate: 320/400 = 0.80 = 80%. Since conditional rates equal the overall rate, there's no association.
There is no association between receiving a promotional offer and customer satisfaction. Both groups have the same 80% satisfaction rate, which also equals the overall satisfaction rate of 80%. The promotional offers did not impact satisfaction levels.
• Association Test: Compare conditional frequencies to marginal frequencies
• No Association: Conditional frequencies approximately equal marginal frequencies
• Independence: When conditional probabilities equal marginal probabilities
Two-Way Table: A table that displays the frequency distribution of two categorical variables, allowing for analysis of their relationship.
Joint Frequency: The count in each individual cell of the table representing occurrences of both categories simultaneously.
Marginal Frequency: The totals for each row and column, found in the margins of the table, representing occurrences of one category regardless of the other.
Joint Relative Frequency: The proportion of the grand total represented by each individual cell (Joint Frequency ÷ Grand Total).
Marginal Relative Frequency: The proportion of the grand total represented by each row or column total (Marginal Frequency ÷ Grand Total).
Conditional Relative Frequency: The proportion of one category within a specific subset of another category (Joint Frequency ÷ Relevant Marginal Frequency).
Statistical Association: A relationship between two variables where knowing the value of one variable provides information about the value of the other variable.
- Table Construction: Organize categorical data into rows and columns
- Table Completion: Fill in all cells and calculate row/column totals
- Joint Frequencies: Record the count in each cell
- Marginal Frequencies: Calculate row and column totals
- Relative Frequencies: Convert to proportions or percentages
- Conditional Analysis: Calculate conditional frequencies to detect associations
- Interpretation: Draw conclusions about relationships between variables
Frequency Calculations:
• Joint Relative Frequency: Cell Value ÷ Grand Total
• Marginal Relative Frequency: Row/Column Total ÷ Grand Total
• Conditional Relative Frequency: Joint Frequency ÷ Relevant Marginal Frequency
Association Detection:
• Strong Association: Large differences in conditional frequencies between groups
• No Association: Conditional frequencies approximately equal between groups
• Independence: Conditional probabilities equal marginal probabilities
Simple Table: 2×2 table with basic counts
Joint Relative Freq: Each cell ÷ total (shows individual probability)
Marginal Relative Freq: Row/col totals ÷ total (shows category probability)
Conditional Freq: Joint ÷ marginal (shows probability given condition)
Association Test: Compare conditional frequencies between groups
Real-World Application: Use for decision-making and predictions
Independence: Two variables are independent if P(A|B) = P(A), meaning knowing B provides no information about A.
Chi-Square Test: A statistical test to determine if observed frequencies significantly differ from expected frequencies under independence.
Expected Frequencies: Under independence assumption, expected frequency = (row total × column total) ÷ grand total.
- Data Organization: Systematically arrange categorical data in table format
- Pattern Recognition: Look for trends and potential associations
- Statistical Analysis: Calculate various types of relative frequencies
- Hypothesis Testing: Assess strength of associations
- Conclusion Drawing: Make evidence-based statements about relationships
- Application: Use findings for predictions and decision-making
• Accurate Data Entry: Ensure all table values are correct and complete
• Proper Calculations: Use correct denominators for different types of relative frequencies
• Meaningful Interpretation: Connect statistical results to real-world context
• Appropriate Conclusions: Avoid overstating the strength of associations
• Context Consideration: Account for relevant external factors
• Clear Communication: Present findings in accessible language
Questions & Answers
Question: What's the difference between joint, marginal, and conditional relative frequencies? How do I know which one to use?
Answer: Each type serves a different purpose:
- Joint Relative Frequency: Shows the probability of both events happening together (cell ÷ grand total). Use when you want to know the likelihood of a specific combination.
- Marginal Relative Frequency: Shows the probability of a single event regardless of the other variable (row/column total ÷ grand total). Use when you want the overall probability of one category.
- Conditional Relative Frequency: Shows the probability of one event given that another has occurred (joint ÷ marginal). Use when you want to know how one variable affects another.
Use conditional frequencies when exploring relationships or associations between variables. Use marginal frequencies for overall probabilities. Use joint frequencies to understand specific combinations.
Question: How do I determine if there's an association between two variables using a two-way table?
Answer: To determine association:
- Calculate conditional relative frequencies for each category of one variable
- Compare these conditional frequencies across the categories
- If the conditional frequencies are significantly different, there's an association
- If the conditional frequencies are approximately equal, there's no association
For example, if 80% of Group A has Characteristic X but only 30% of Group B has Characteristic X, there's likely an association. The larger the difference in conditional frequencies, the stronger the association.
Remember: association does not imply causation!
Question: Why do we need to learn about relative frequencies when we already have regular frequencies?
Answer: Relative frequencies are essential because they allow for fair comparisons:
- Standardization: Convert raw counts to proportions that account for different group sizes
- Comparison: Compare groups of different sizes meaningfully
- Probability: Relative frequencies approximate probabilities in the population
- Patterns: Easier to spot trends and relationships in the data
For example, if Group A has 10 successes out of 20 trials and Group B has 100 successes out of 200 trials, both have the same success rate (50%) despite very different raw numbers. Relative frequencies reveal this similarity.
Regular frequencies tell us counts, but relative frequencies tell us rates and probabilities.