Section 1.7: Math, Statistics, and Data Interpretation

Descriptive Statistics and Summary Measures

Quantitative Job Knowledge questions on the FSOT test your ability to apply statistical and mathematical concepts to policy analysis.

Measures of Central Tendency

Analysts use three primary measures of center to summarize data:

• Mean (arithmetic average): The sum of all values divided by the number of observations (n). It is best suited for symmetric distributions without extreme values.
• Median: The middle value when the data are ordered from lowest to highest (or the average of the two middle values if n is even). The median is robust to outliers, making it the preferred measure for skewed distributions (e.g., host-country household income or visa processing times).
• Mode: The most frequently occurring value in a dataset, useful for categorical data.

Weighted averages are required when combining groups of different sizes. If a country has two provinces, Province A (population 10 million) with an unemployment rate of 5% and Province B (population 2 million) with an unemployment rate of 11%, the weighted average is:

Weighted Average = ((10 * 5%) + (2 * 11%)) / (10 + 2) = (50% + 22%) / 12 = 6%

Simply averaging 5% and 11% to get 8% is incorrect because it treats both provinces equally despite their size difference.

Spread, Percentages, and Ratios

The range is the difference between the maximum and minimum values. Understanding spread is vital because two datasets can share the same mean while representing vastly different realities (e.g., stable daily visa issuance vs. highly volatile fluctuations).

Calculating percent change requires dividing the difference by the baseline (old) value:

Percent Change = ((New Value - Old Value) / Old Value) * 100%

Trap: Confusing a percent change with a change in percentage points. If a country's tariff rate increases from 10% to 12%, the rate has increased by 2 percentage points, but the relative percent increase in the tariff is 20% (2 / 10).

Additionally, successive changes do not cancel out: a 10% budget increase followed by a 10% cut results in a net 1% loss (100 * 1.10 = 110, and 110 * 0.90 = 99).

Elementary Probability and Decision Theory

Probability measures the likelihood of an event occurring, ranging from 0 (impossible) to 1 (certain).

• Complement Rule: The probability of an event not occurring is 1 - P(A).
• Independent Events: If two events do not affect each other, the probability of both occurring is the product of their individual probabilities: P(A and B) = P(A) * P(B).
• Mutually Exclusive Events: If two events cannot happen simultaneously, the probability of either occurring is the sum of their individual probabilities: P(A or B) = P(A) + P(B).

Expected Value is the weighted average of all possible outcomes, weighted by their probabilities:

Expected Value = Sum of (Outcome * Probability)

For example, if a developmental project has a 70% chance of yielding a $10 million return and a 30% chance of a $5 million loss, the expected value is:

Expected Value = (10 * 0.70) + (-5 * 0.30) = 7 - 1.5 = $5.5 million

Data Interpretation, Graph Analysis, and Common Biases

Diplomats must review charts and tables efficiently without falling for visual or statistical manipulation.

Fixed Pre-Read Checklist

Before attempting to answer a graphic-based question, complete this sequence:

1. Verify Title & Subtitle: Confirm the exact scope of the data.
2. Examine Units & Scales: Look for notations like "in thousands" or "per 100,000." Note if the y-axis is truncated (does not start at zero), which exaggerates small changes.
3. Check the Legend: Identify which line or bar represents which variable.
4. Identify Time Horizon: Confirm whether dates refer to calendar years, fiscal years, or quarters.

Graphic Types and Misuses

Different visuals serve specific functions:

• Line Charts: Excellent for tracking trends over time. Do not confuse a level (high point) with a rate of change (slope).
• Bar Charts: Best for comparing distinct categories.
• Pie Charts: Show shares of a whole at a single point in time. Never compare shares across different pies without checking if the absolute totals differ.
• Scatterplots: Display the relationship between two variables.

Index Numbers

Index numbers compare values relative to a base year set at 100. If an import price index is 100 in 2020 and 120 in 2024, import prices rose 20% over that period. If the index rises to 132 in 2025, the increase from 2024 to 2025 is 10% ((132 - 120) / 120), not 12%.

Statistical Fallacies and Biases

• Correlation vs. Causation: Co-movement between two variables (e.g., GDP growth and internet penetration) does not prove one caused the other. Both could be driven by a third factor, like regulatory reform.
• Selection Bias: Drawing conclusions from an unrepresentative sample, such as surveying only urban businesses to assess national economic sentiment.
• Base-Year Manipulation: Distorting trends by starting a chart in an unusually high or low year.