เครื่องคำนวณส่วนเบี่ยงเบนมาตรฐาน

วิธีใช้เครื่องคิดเลขนี้

1. ป้อนNumbers (comma-separated)
2. คลิกปุ่มคำนวณ
3. อ่านผลลัพธ์ที่แสดงด้านล่างเครื่องคิดเลข

What Is Standard Deviation and Why Does It Matter?

Standard deviation measures how spread out your data is around the mean (average). A small standard deviation means values cluster tightly around the mean; a large standard deviation means values are widely scattered.

Two datasets can have the same average but completely different distributions — standard deviation captures that difference:

• Dataset A: {9, 10, 10, 11, 10} — Mean = 10, SD ≈ 0.63 (tight cluster)
• Dataset B: {2, 5, 10, 15, 18} — Mean = 10, SD ≈ 5.83 (widely spread)

Both have a mean of 10, but Dataset B is nearly 10× more variable. Standard deviation makes this visible.

Standard deviation is denoted σ (sigma) for a population and s for a sample. It is the square root of variance, expressed in the same units as the original data — making it more interpretable than variance alone.

Applications span almost every field: quality control (are manufactured parts consistently within tolerance?), finance (investment risk = return volatility), medicine (is a patient's reading within 2 SD of normal?), education (how are test scores distributed?), and sports analytics (how consistent is an athlete's performance?).

Population vs Sample Standard Deviation

The most important choice when calculating standard deviation is whether you're working with a population (all possible data points) or a sample (a subset). This determines which formula to use and affects the result.

Population standard deviation (σ): Use when you have data for the entire group you're studying. Formula: σ = √[Σ(xᵢ − μ)² / N]

Where: μ = population mean, N = number of values, Σ = sum of all values.

Sample standard deviation (s): Use when your data is a sample drawn from a larger population. Formula: s = √[Σ(xᵢ − x̄)² / (n−1)]

Where: x̄ = sample mean, n = number of values in sample, (n−1) = Bessel's correction.

Bessel's correction divides by (n−1) instead of n because samples tend to underestimate the true population variance — particularly for small samples. Using (n−1) provides an unbiased estimator of the population variance.

Which to use?

• Population SD: You have data for all students in a specific class; all the test scores from one specific exam; all employees at a single company.
• Sample SD: You surveyed 500 Americans about income (inferring to all Americans); you measured 30 widgets from a production run (inferring to all widgets); any scientific study with a sample.

Step-by-Step Standard Deviation Calculation

Let's work through a complete example with real numbers:

Dataset: Test scores of 6 students: {72, 85, 91, 68, 79, 88}

Step 1 — Find the mean: (72 + 85 + 91 + 68 + 79 + 88) / 6 = 483 / 6 = 80.5

Step 2 — Find each deviation from mean and square it:

Step 3 — Calculate variance: Sample variance (n−1) = 417.50 / 5 = 83.50

Step 4 — Take square root for standard deviation: s = √83.50 ≈ 9.14

Interpretation: Most scores fall within about 9.14 points of the 80.5 mean. Approximately 68% of scores would be expected between 71.4 and 89.6 (mean ± 1 SD) if this were a normally distributed population.

The Empirical Rule and Normal Distribution

For data that follows a normal distribution (bell curve), the Empirical Rule (68-95-99.7 rule) tells you exactly how many values fall within each standard deviation range:

The classic application is IQ scores: mean = 100, SD = 15. An IQ of 130 is 2 SDs above the mean — only about 2.3% of people score that high. An IQ of 145 is 3 SDs above the mean — about 0.13% of people (roughly 1 in 750).

In quality control, the Six Sigma standard requires processes to have fewer than 3.4 defects per million opportunities — equivalent to keeping variation within ±6 standard deviations from the target, leaving only 0.00034% defect rate. This is the statistical foundation of Six Sigma manufacturing quality programs.

Not all data is normally distributed. Income distributions are right-skewed (a few very high earners stretch the right tail). In such cases, the median and interquartile range may be more informative than mean and standard deviation.

Other Statistical Measures: Mean, Median, Variance, and More

Standard deviation is most meaningful alongside other descriptive statistics. Here's how they work together:

• Mean (arithmetic average): Sum of all values ÷ count. Sensitive to outliers — one extreme value can significantly shift the mean.
• Median: The middle value when data is sorted. More robust to outliers than the mean. For {1, 2, 3, 4, 100}: mean = 22, median = 3.
• Mode: The most frequently occurring value. Useful for categorical data; a dataset can have multiple modes or none.
• Range: Maximum − minimum. Simple but sensitive to outliers; doesn't describe the distribution shape.
• Variance (σ² or s²): The square of standard deviation. Useful mathematically but harder to interpret since it's in squared units. Example: if heights are in centimeters, variance is in cm² — which has no physical meaning.
• Coefficient of Variation (CV): (Standard deviation / mean) × 100%. Allows comparing variability across datasets with different means. A CV of 10% means the SD is 10% of the mean — useful in finance and biology.
• Standard Error of the Mean (SEM): SD ÷ √n. Measures the precision of the sample mean as an estimate of the population mean. As sample size grows, SEM shrinks — larger samples give more precise estimates.

Standard Deviation in Finance, Science, and Sports

Standard deviation has specific, practical interpretations across different fields:

Finance — Measuring investment risk: In finance, standard deviation of returns = volatility = risk. A stock returning 10% annually with SD of 15% has a 68% probability of returning between −5% and +25% in any given year. The S&P 500 historically has an annual SD of about 15–20%. Bond portfolios typically have SD of 3–7%. Risk-adjusted performance (Sharpe Ratio) = (return − risk-free rate) / SD — the higher, the better.

Science — Quality control and measurement: Laboratory instruments report measurements as mean ± SD. A thermometer reading 37.2 ± 0.3°C means the measurement is within 0.3°C of the true value with 68% confidence. In clinical trials, statistical significance is typically defined as the treatment effect being more than 2 SDs from the control group mean (p < 0.05).

Sports analytics: Player consistency is quantified with SD. A basketball player averaging 25 points per game with SD of 3 is more reliable than one averaging 25 with SD of 10. Weather forecasting uses ensemble models where the SD of temperature predictions indicates confidence — a narrow SD means forecasters agree; a wide SD means high uncertainty.

Education: Z-scores express how many standard deviations a student's score is from the class mean: Z = (score − mean) / SD. A Z-score of +2 means scoring 2 SDs above the mean — better than approximately 97.7% of students. Standardized tests like the SAT are designed so scores follow a roughly normal distribution, enabling these percentile comparisons.