12.3 DS Applied to Overlapping Sets, Venn Diagrams, and Probability
Key Takeaways
- In overlapping set problems with two binary characteristics, organizing data into a 2x2 double-set matrix reveals that out of 9 total cells, exactly 4 independent cell values are required to completely determine the entire matrix.
- To solve for a single target cell or marginal total in a double-set matrix, determining all individual cells is unnecessary if a given statement directly constrains the target cell via a linear aggregate.
- In three-set Venn diagrams, the master equation Total = A + B + C - (sum of 2-set overlaps) + (all three) + Neither connects the aggregate groups; distinguish carefully between 'both A and B' (which includes the center) and 'only A and B' (which excludes the center).
- Two events A and B are statistically independent if and only if P(A ∩ B) = P(A) × P(B); in Data Sufficiency, an assertion of independence provides an independent non-linear equation that eliminates one degree of freedom.
- Without an explicit statement of independence or mutual exclusivity, knowing individual probabilities P(A) and P(B) is strictly insufficient to determine joint probability P(A ∩ B) or union probability P(A ∪ B).
12.3 DS Applied to Overlapping Sets, Venn Diagrams, and Probability
Quick Summary: Overlapping sets and probability questions on GMAT Focus Data Sufficiency evaluate your mastery of mathematical systems and constraint satisfaction. For two overlapping categories, the 2x2 Double-Set Matrix is the ultimate operational tool: four mutually exclusive internal cells sum to four marginal totals and one grand total. While the complete matrix has 4 degrees of freedom, targeted questions often require only 1 or 2 specific constraints. For three-set problems, distinguish between pairwise overlaps ($AB + BC + AC$) and exact two-set regions (which strip out the triple intersection). In probability, sufficiency hinges on structural event definitions: knowing $P(A)$ and $P(B)$ can never determine $P(A \cap B)$ or $P(A \cup B)$ unless the prompt or statements explicitly confirm that the events are independent ($P(A \cap B) = P(A)P(B)$) or mutually exclusive ($P(A \cap B) = 0$).
The Double-Set Matrix Framework in Data Sufficiency
When a problem classifies a population across two binary characteristics (e.g., Male vs. Female, Enrolled vs. Not Enrolled, Product A vs. Not Product A), drawing Venn circles is prone to spatial confusion. The Double-Set Matrix organizes the sample space into a 9-cell system of linear equations:
| Characteristic $B$ | Not Characteristic $B$ ($\sim B$) | Total | |
|---|---|---|---|
| Characteristic $A$ | $A \cap B$ | $A \cap \sim B$ | $\mathbf{Total}(A)$ |
| Not Characteristic $A$ ($\sim A$) | $\sim A \cap B$ | $\sim A \cap \sim B$ | $\mathbf{Total}(\sim A)$ |
| Total | $\mathbf{Total}(B)$ | $\mathbf{Total}(\sim B)$ | $\mathbf{Grand~Total}~N$ |
Degrees of Freedom and Sufficiency Mechanics
The 9-cell matrix contains:
- 4 mutually exclusive interior cells: $(A \cap B)$, $(A \cap \sim B)$, $(\sim A \cap B)$, and $(\sim A \cap \sim B)$.
- 4 marginal totals: $\text{Total}(A)$, $\text{Total}(\sim A)$, $\text{Total}(B)$, and $\text{Total}(\sim B)$.
- 1 grand total: $N$.
Because row sums and column sums create 5 linear dependencies ($r_1 + r_2 = \text{row sum}$, etc.), the entire 9-cell system has exactly 4 degrees of freedom (or 3 degrees of freedom if the Grand Total $N$ is already fixed in the question stem).
The Targeted Sufficiency Shortcut:
If the question stem asks for a specific cell (e.g., "How many executives belong to both Group A and Group B?"), you do NOT need to resolve all 4 degrees of freedom! You only need an equation that links the target cell directly to known totals via the master inclusion-exclusion identity:
If the question stem provides $N$, $\text{Total}(A)$, and $\text{Total}(B)$, then finding $(A \cap B)$ is mathematically equivalent to finding $(\text{Neither})$. Any statement that provides $(\text{Neither})$ immediately yields sufficiency for $(A \cap B)$!
Three-Set Venn Diagrams: Pairwise Overlaps vs. "Exactly Two"
When three categories overlap ($A, B, C$), the universe decomposes into 8 distinct regions:
- Only $A$, Only $B$, Only $C$ (Exactly 1 set)
- Only $A$ and $B$, Only $B$ and $C$, Only $A$ and $C$ (Exactly 2 sets)
- All three sets ($A \cap B \cap C$, Exactly 3 sets)
- Neither set (${0}$, None)
[ Three-Set Decomposition ]
┌───────────────────────────┴───────────────────────────┐
▼ ▼
[ Master Inclusion-Exclusion ] [ Partition Formula ]
Total = A + B + C Total = (Exactly 1)
- (AB + BC + AC) + (Exactly 2)
+ (A ∩ B ∩ C) + None + (Exactly 3) + None
The Two Master Equations
\textbf{Equation 1 (Gross Intersections):} \quad N &= A + B + C - (AB + BC + AC) + (A \cap B \cap C) + \text{None} \\[6pt] \textbf{Equation 2 (Disjoint Partitions):} \quad N &= (\text{Exactly 1}) + (\text{Exactly 2}) + (\text{Exactly 3}) + \text{None} \end{aligned}$$ ### The Definitive GMAT Sufficiency Trap: Pairwise vs. Disjoint Overlaps Notice the difference between the two overlap expressions: $$(AB + BC + AC) = (\text{Exactly 2}) + 3 \times (\text{Exactly 3})$$ - The term $(AB)$ represents all elements common to $A$ and $B$, **including those that also belong to $C$**. - The term *"Only $A$ and $B$"* excludes the triple intersection. - If a statement says *"15 students study both French and Spanish"*, that is a gross intersection ($AB = 15$). If it says *"15 students study French and Spanish but NOT German"*, that is a disjoint component of (Exactly 2). - For Data Sufficiency, to determine $(A \cap B \cap C)$ from Equation 1, you must know $(A + B + C)$, $\text{None}$, and the gross sum $(AB + BC + AC)$. Knowing only one pairwise overlap (e.g., just $AB$) leaves the other two pairwise terms unconstrained, producing an indeterminate system. --- ## Probability Sufficiency: Independence and Conditional Constraints Probability questions on Data Sufficiency test whether you can recognize when an event space is constrained versus when it contains unconstrained correlation. ### The Addition Rule in DS $$P(A \cup B) = P(A) + P(B) - P(A \cap B)$$ This equation contains 4 variables. Knowing any three allows you to solve for the fourth. If a question stem asks for $P(A \cup B)$ (the probability that at least one event occurs), knowing $P(A)$ and $P(B)$ alone is **NOT sufficient** because the joint probability $P(A \cap B)$ can range anywhere within the Fréchet bounds: $$\max(0, P(A) + P(B) - 1) \le P(A \cap B) \le \min(P(A), P(B))$$ ### The Power of the Independence Assertion If the prompt or a statement explicitly asserts that **Events $A$ and $B$ are independent**, a non-linear algebraic relationship is established: $$P(A \cap B) = P(A) \times P(B)$$ This eliminates one degree of freedom entirely! When events are independent: $$P(A \cup B) = P(A) + P(B) - P(A)P(B) = 1 - (1 - P(A))(1 - P(B))$$ $$\text{And} \quad P(\sim A \cap \sim B) = (1 - P(A))(1 - P(B))$$ *Sufficiency Rule:* If $A$ and $B$ are independent, knowing $P(A)$ and $P(B)$ is **ALONE sufficient** to compute $P(A \cap B)$, $P(A \cup B)$, $P(A \mid B)$, and $P(\text{Neither})$! ### Mutually Exclusive Events If events $A$ and $B$ are **mutually exclusive** (disjoint), they cannot occur simultaneously: $P(A \cap B) = 0$. In this case, $P(A \cup B) = P(A) + P(B)$. Again, this eliminates the joint overlap variable. ### Conditional Probability Sufficiency $$P(A \mid B) = \frac{P(A \cap B)}{P(B)} \implies P(A \cap B) = P(A \mid B) \times P(B)$$ If a statement provides the conditional probability $P(A \mid B)$ and the marginal probability $P(B)$, their product immediately fixes the intersection $P(A \cap B)$, achieving sufficiency for intersection or union questions. --- ## Complete Worked DS Scenarios ### Worked Example 1: 2x2 Matrix Sufficiency for Corporate Staffing > **Question Stem:** > In an international bank branch with 150 total employees, each employee is either an account manager or a compliance officer, and each employee works either primarily remotely or primarily on-site. How many compliance officers at the branch work primarily on-site? > > **Statement (1):** At the branch, 65 employees are account managers who work primarily remotely, and 45 employees are compliance officers. > > **Statement (2):** At the branch, 80 employees work primarily remotely. #### Analytical Deconstruction Set up the 2x2 matrix with rows representing Job Role (Account Manager $A$, Compliance Officer $C$) and columns representing Work Location (Remote $R$, On-Site $O$). $$\begin{array}{|c|c|c|c|} \hline \textbf{Role / Location} & \textbf{Remote } (R) & \textbf{On-Site } (O) & \textbf{Total} \\ \hline \textbf{Account Manager } (A) & A_R & A_O & A_{\text{total}} \\ \hline \textbf{Compliance Officer } (C) & C_R & \mathbf{C_O \text{ (Target)}} & C_{\text{total}} \\ \hline \textbf{Total} & R_{\text{total}} & O_{\text{total}} & N = 150 \\ \hline \end{array}$$ - **Evaluating Statement (1) Alone:** Statement (1) gives $A_R = 65$ and $C_{\text{total}} = 45$. Since $N = 150$ and $C_{\text{total}} = 45$, we can deduce $A_{\text{total}} = 150 - 45 = 105$. Since $A_{\text{total}} = A_R + A_O$, we have $105 = 65 + A_O \implies A_O = 40$. However, for compliance officers, we only know $C_R + C_O = 45$. The target $C_O$ can be any integer from 0 to 45 (for example, $C_O = 10$ and $C_R = 35$, or $C_O = 25$ and $C_R = 20$). Statement (1) alone is **NOT sufficient**. - **Evaluating Statement (2) Alone:** Statement (2) gives $R_{\text{total}} = 80$. This allows us to find $O_{\text{total}} = 150 - 80 = 70$. However, $O_{\text{total}} = A_O + C_O = 70$. Without information dividing the on-site employees between account managers and compliance officers, $C_O$ is undetermined. Statement (2) alone is **NOT sufficient**. - **Evaluating Statements (1) and (2) Together:** From Statement (1), we found $A_O = 40$. From Statement (2), we found $O_{\text{total}} = 70$. Because the on-site column must sum to the total on-site count: $$A_O + C_O = O_{\text{total}} \implies 40 + C_O = 70 \implies C_O = 30$$ The target cell $C_O$ is uniquely fixed at 30. **Conclusion:** Both statements together are sufficient, but neither statement alone is sufficient. --- ### Worked Example 2: Probability and Independent Systems > **Question Stem:** > A quality-control testing station evaluates microchips for two independent defects: solder bridging (Defect A) and silicon cracking (Defect B). What is the probability that a randomly selected microchip from the production line has neither defect? > > **Statement (1):** The probability that a randomly selected microchip has Defect A is 0.08, and the probability that it has Defect B is 0.05. > > **Statement (2):** The probability that a randomly selected microchip has both Defect A and Defect B is 0.004. #### Analytical Deconstruction The prompt establishes a foundational physical constraint: **Defect A and Defect B are INDEPENDENT**. Therefore, by definition of statistical independence: $$P(A \cap B) = P(A) \times P(B)$$ $$P(\text{neither}) = P(\sim A \cap \sim B) = (1 - P(A)) \times (1 - P(B)) = 1 - [P(A) + P(B) - P(A)P(B)]$$ - **Evaluating Statement (1) Alone:** Statement (1) gives $P(A) = 0.08$ and $P(B) = 0.05$. Substitute directly into the independent probability identity: $$P(\text{neither}) = (1 - 0.08)(1 - 0.05) = (0.92)(0.95) = 0.874$$ This provides a single, unique numerical value. Statement (1) alone is **SUFFICIENT**. - **Evaluating Statement (2) Alone:** Statement (2) gives $P(A \cap B) = 0.004$. Since $A$ and $B$ are independent, $P(A) \times P(B) = 0.004$. Infinitely many pairs of probabilities satisfy this equation: - If $P(A) = 0.08$ and $P(B) = 0.05$, then $P(\text{neither}) = (0.92)(0.95) = 0.874$. - If $P(A) = 0.04$ and $P(B) = 0.10$, then $P(\text{neither}) = (0.96)(0.90) = 0.864$. Because different probabilities yield different values for $P(\text{neither})$, Statement (2) alone is **NOT sufficient**. **Conclusion:** Statement (1) alone is sufficient, but statement (2) alone is not sufficient. --- ## High-Frequency Traps in Overlapping Sets and Probability DS 1. **The Phantom Independence Assumption:** Assuming two real-world events are independent when the prompt does not state they are independent. Unless the prompt explicitly confirms independence, you must assume $P(A \cap B)$ is an independent unknown variable. 2. **Ignoring the "Neither" Group:** In overlapping set problems, candidates often set $\text{Total} = A + B - (A \cap B)$, completely forgetting that some individuals may belong to neither category. Never omit the $\text{Neither}$ cell in the 2x2 matrix. 3. **Confusing Overlap Classifications in 3-Set Venn Diagrams:** Treating the phrase *"students who like both Math and Physics"* as meaning students who like *only* Math and Physics. On the GMAT, "both A and B" includes those who like all three unless the word "only" is explicitly specified. 4. **Over-solving the 2x2 Matrix:** Spending two minutes trying to compute all 9 cells of a double-set matrix when a single column or row relationship already answers the target question.In a survey of 200 corporate executives, each executive was asked whether their company offers flexible work hours and whether their company offers subsidized gym memberships. How many of the executives surveyed reported that their company offers both flexible work hours and subsidized gym memberships? Statement (1): 130 executives reported that their company offers flexible work hours, and 90 executives reported that their company offers subsidized gym memberships. Statement (2): 30 executives reported that their company offers neither flexible work hours nor subsidized gym memberships.
In an international academy of 100 students, each student studies at least one of three foreign languages: Spanish, French, or German. How many students study all three languages? Statement (1): Exactly 50 students study Spanish, 40 study French, and 35 study German. Statement (2): Exactly 20 students study both Spanish and French.
A quality-control testing station evaluates microchips for two independent defects: solder bridging (Defect A) and silicon cracking (Defect B). What is the probability that a randomly selected microchip from the production line has neither defect? Statement (1): The probability that a randomly selected microchip has Defect A is 0.08, and the probability that it has Defect B is 0.05. Statement (2): The probability that a randomly selected microchip has at least one of the two defects is greater than 0.10.