In statistics, the difference between biased and unbiased selection is about how representative a sample is of the entire population.
**Biased Selection:**
Imagine you want to understand the average height of all students in a school, but you only measure the height of the basketball team. Since the basketball players are generally taller than average, your sample won’t accurately represent the heights of all students.
**Unbiased Selection:**
Now, if you randomly select students from all grades and classes to measure their heights, you’re more likely to get a sample that represents the entire student body accurately. This method reduces the chance of over-representing any particular group.
In essence, a biased selection skews results because it doesn’t accurately reflect the entire population, while an unbiased selection gives a more accurate picture by representing the population fairly.
The terms `n` and `n-1` come into play when calculating sample statistics, particularly when estimating the population variance or standard deviation from a sample.
**Sample Variance Calculation:**
- **Using `n` (Sample Size):** When calculating the variance of a sample, if you divide the sum of squared deviations from the sample mean by `n`, you get the *sample variance*. This method often underestimates the population variance because it does not account for the fact that the sample mean is an estimate itself, rather than the true population mean.
- **Using `n-1` (Degrees of Freedom):** To correct for this underestimation, we divide by `n-1` instead. This adjustment is known as "Bessel's correction." The resulting value is called the *sample variance*, which provides an unbiased estimate of the population variance.
**Example:**
Suppose you measure the heights of 4 students and get these values: 150 cm, 160 cm, 165 cm, and 170 cm.
1. Calculate the sample mean: `(150 + 160 + 165 + 170) / 4 = 161.25` cm.
2. Find the squared deviations from the mean and sum them up: `(150 - 161.25)^2 + (160 - 161.25)^2 + (165 - 161.25)^2 + (170 - 161.25)^2`.
3. The sum is `126.5625 + 1.5625 + 14.0625 + 76.5625 = 218.75`.
- **Using `n` (4):** Variance = `218.75 / 4 = 54.6875` (this tends to underestimate the true variance of the population).
- **Using `n-1` (3):** Variance = `218.75 / 3 = 72.9167` (this is an unbiased estimate of the population variance).
So, using `n-1` corrects for the bias in the sample variance estimation.
