学生 t 分布

Use the Student t distribution when the population standard deviation is unknown, which is typical in real-world data. It applies to simple random samples from normally distributed populations or when sample size exceeds 30. The t distribution is particularly valuable for small sample sizes, where it provides more accurate confidence intervals than the normal distribution by accounting for greater variability.

The Student t distribution has wider confidence intervals because it exhibits greater variability and always maintains a standard deviation greater than 1. This wider spread reflects the uncertainty introduced when estimating the population standard deviation from sample data. The larger critical values in the t distribution account for this additional variability, making confidence intervals more conservative and reliable for small samples.

As sample size increases, the Student t distribution converges toward the normal distribution. The exact shape depends on degrees of freedom, calculated as n minus 1. With larger degrees of freedom, the t distribution curve becomes thinner in the tails and taller in the center, increasingly resembling the standard normal curve until they are nearly identical at very large sample sizes.

The t score measures how far a value is from its mean, similar to the z score in normal distributions. It is calculated using the sample mean, population mean, sample standard deviation, and sample size. The t score indicates the number of standard errors a sample mean is from the population mean, helping assess whether observed differences are statistically significant.

The Student t distribution has thicker tails because it contains more probability in its tail regions than the standard normal distribution. This reflects the greater spread and variability inherent in the t distribution, which accounts for uncertainty when estimating population parameters from small samples. The thicker tails provide more conservative critical values for hypothesis testing and confidence interval construction.

The Student t distribution was developed by William S. Gosset (1876–1937), a statistician at the Guinness brewery in Dublin, Ireland. Gosset created this distribution to solve the problem of inaccurate confidence intervals when sample sizes were small. He published his work under the pen name "Student," which is why the distribution bears this name today.

The margin of error in the Student t distribution is evaluated using a specific formula that incorporates the t critical value, sample standard deviation, and sample size. This calculated margin of error is then used to establish the confidence interval limits around the sample mean. The formula accounts for the increased variability present in small samples, ensuring more accurate interval estimates.