Eighth Grade Mathematics: Calculating Conditional Relative Frequencies
While marginal and joint relative frequencies compare counts to the grand total of all surveyed individuals, the most powerful insights in statistics arise when you restrict your focus to a single specific subpopulation. For example, rather than asking what percentage of all students play video games, you might ask: of the students who play video games, what percentage earn honor roll grades? A proportion calculated within a restricted row or column category is called a conditional relative frequency. In this chapter, you will master calculating row and column conditional relative frequencies, understand how the conditioning denominator works, and use conditions to analyze behavior.
The Definition of Conditional Relative Frequency
A conditional relative frequency is the ratio of a joint frequency (an inner cell) to a specific marginal total (either a row total or a column total), rather than the grand total:
Conditional Relative Frequency Formulas:
1. Row Conditional Relative Frequency (Conditioned on a ROW):
Row Conditional Frequency = (Cell Count) / (ROW TOTAL)
"Given that an individual is in this ROW, what is the probability of being in this column?"
2. Column Conditional Relative Frequency (Conditioned on a COLUMN):
Column Conditional Frequency = (Cell Count) / (COLUMN TOTAL)
"Given that an individual is in this COLUMN, what is the probability of being in this row?"
The critical phrase to recognize in verbal problems is given that, of the, or among. The group specified after given that or of establishes the restricted condition and dictates the number that MUST be placed in the denominator.
Calculating Row Conditional Relative Frequencies
When you condition by row, each individual row becomes its own isolated 100% universe:
Raw Table: Music Preference by Age Group
Prefers Pop Prefers Rock Row Total
Teens (13-19) 75 25 100
Adults (30-49) 30 70 100
Column Total 105 95 200
Row Conditional Relative Frequency Table (Dividing each cell by its ROW TOTAL):
Prefers Pop Prefers Rock Total
Teens 75/100 = 0.75 (75%) 25/100 = 0.25 (25%) 100/100 = 1.00 (100%)
Adults 30/100 = 0.30 (30%) 70/100 = 0.70 (70%) 100/100 = 1.00 (100%)
In this row table, every horizontal row sums to 1.00 (100%). It answers questions about age groups:
- Of the teens, what percent prefer pop? Answer: 75%.
- Of the adults, what percent prefer pop? Answer: 30%.
Comparing 75% to 30% immediately reveals a massive behavioral difference between the age groups that was not as obvious from the raw counts.
Calculating Column Conditional Relative Frequencies
When you condition by column, each vertical column becomes its own isolated 100% universe, dividing each inner cell by the column total at the bottom:
Column Conditional Relative Frequency Table (Dividing each cell by its COLUMN TOTAL):
Prefers Pop Prefers Rock
Teens 75/105 = 0.714 (71.4%) 25/95 = 0.263 (26.3%)
Adults 30/105 = 0.286 (28.6%) 70/95 = 0.737 (73.7%)
Total 105/105 = 1.00 (100%) 95/95 = 1.00 (100%)
In this column table, each vertical column sums to 1.00 (100%). It answers questions about musical tastes:
- Of the pop listeners, what percent are teens? Answer: 71.4%.
- Of the rock listeners, what percent are adults? Answer: 73.7%.
Notice that the question of the teens, what percent prefer pop (75%) is fundamentally different from of the pop listeners, what percent are teens (71.4%). The denominator changes completely based on the conditioning group.
Chapter Exercises
Exercise 1: Using the raw table tracking 150 students: 40 boys play soccer, 20 boys do not play soccer (total 60 boys); 30 girls play soccer, 60 girls do not play soccer (total 90 girls). Calculate the conditional relative frequency of playing soccer given that a student is a boy, and the conditional relative frequency of playing soccer given that a student is a girl. Compare the two proportions.
Exercise 2: Using the same table from Exercise 1, calculate the conditional relative frequency of being a boy given that a student plays soccer. Contrast this calculation with the calculation performed in Exercise 1.
Exercise 3: Explain in paragraph form why you cannot simply add the percentages in a conditional relative frequency table across both rows and columns to equal 100%, discussing how the reference whole changes.
Solutions and Explanations
Solution 1: To find the conditional relative frequency of playing soccer given that a student is a boy, restrict the analysis to the row of boys. The total number of boys is 60, and the number of boys who play soccer is 40. The conditional relative frequency is 40 / 60 = 2/3 = 0.667, or approximately 66.7 percent. To find the conditional relative frequency of playing soccer given that a student is a girl, restrict the analysis to the row of girls. The total number of girls is 90, and the number of girls who play soccer is 30. The conditional relative frequency is 30 / 90 = 1/3 = 0.333, or approximately 33.3 percent. Comparing the two conditional proportions: boys in this survey are twice as likely to play soccer as girls (66.7% versus 33.3%).
Solution 2: The question asks for the conditional relative frequency of being a boy given that the student plays soccer. Here, the condition is plays soccer, meaning the analysis is restricted to the column of soccer players. The total number of soccer players is 40 boys + 30 girls = 70 soccer players. The number of boys in this group is 40. The conditional relative frequency is 40 / 70 = 4/7 = 0.571, or approximately 57.1 percent. This contrasts sharply with Exercise 1: in Exercise 1, the condition was being a boy, which used the total of boys (60) in the denominator to yield 66.7%. In Exercise 2, the condition is playing soccer, which uses the total of soccer players (70) in the denominator to yield 57.1%. Swapping the condition changes the reference denominator and answers a completely different statistical question.
Solution 3: In a conditional relative frequency table, you cannot add percentages across both rows and columns simultaneously to equal 100% because the reference whole (the base denominator) changes depending on which direction you conditioned. If a table is constructed using row conditional relative frequencies, each individual row represents a completely separate subpopulation whose cells were divided by that specific row's marginal total. Consequently, each horizontal row independently sums to 1.00 (100%). However, the columns cannot sum to 100% because the numbers in a column were derived from different denominators representing different group sizes. Adding percentages that have different base wholes is mathematically invalid—just as adding 50% of a small apple to 50% of a giant watermelon does not equal 100% of any single fruit. A conditional table is strictly normalized in one direction: either every row sums to 100%, or every column sums to 100%, but never both.