3.2. Deviation from the Mean

Each data point deviates from the mean:

Deviation = 𝑥 −𝑥‾

In this chunk, we focus on the concept of deviation, which is a measure of how much a specific data point differs from the average (the mean) of all the data points. If we have a data set and we calculate the mean, the deviation for each data point is found by subtracting the mean from that data point. This tells us whether that specific data value is above or below the average. The formula presented shows that this calculation can result in both positive and negative values, indicating whether the data point is above the mean (positive deviation) or below the mean (negative deviation).

Imagine you are comparing the heights of your friends. If the average height is 160 cm, and one of your friends is 170 cm tall, their deviation from the mean height is +10 cm, indicating they are taller than average. Conversely, if another friend is 150 cm tall, their deviation is -10 cm, indicating they are shorter than average.

Variance is the average of the squared deviations from the mean.

• For a population: ∑(𝑥 −𝜇)² / 𝑁

• For a sample: ∑(𝑥 −𝑥‾)² / (𝑛−1)

Variance represents how spread out the data points are relative to the mean. To calculate variance, we take each deviation from the mean, square it (to eliminate negative values), then find the average of these squared deviations. The formula can differ depending on whether we are looking at an entire population or just a sample. The population formula divides by the total number of data points (N), while the sample formula divides by one less than the number of data points (n-1) to account for sample variability.

If we were to track the daily temperatures in your city over a week, we might find that most temperatures hover around 20°C, but occasionally it spikes to 30°C or drops to 10°C. By using the variance calculation, we can see how much those occasional extreme temperatures deviate from the average—showing us that even though most days are similar, there is significant variation at times.

Standard deviation is the square root of the variance.

• For a population: ∑(𝑥 −𝜇)² / 𝑁

• For a sample: ∑(𝑥 −𝑥‾)² / (𝑛−1)

The standard deviation is a key statistic because it puts the measure of spread back into the original units of the data, making it easier to interpret. By taking the square root of the variance, we obtain a measure that tells us, on average, how far each data point is from the mean. This helps in understanding the data distribution more intuitively.

Consider you are evaluating the performance of students in a math test. If the standard deviation is low, say close to 1, it means most students scored around the average score, say 75%. However, if the standard deviation is high, such as 10, it indicates that scores vary widely, meaning some students scored very high while others scored very low—showing a diverse level of understanding in the class.

• Avoids negatives: Without squaring, the sum of deviations would always be zero. • Penalizes large deviations: Squaring gives more weight to larger differences.

Squaring the differences is essential to ensuring that all deviations contribute positively to the overall calculation. If we didn't square the differences, deviations above and below the mean would cancel each other out, leading to a misleading result (a deviation total of zero). Additionally, squaring gives significantly larger weights to more substantial deviations, thus emphasizing their impact on the variance and standard deviation metrics.

Think about your savings over a month. If you usually save around $100 but one month you save$200, while another month you save $0, simply averaging those would not give a true picture of your saving behavior. If we square the differences from your average savings, we can better capture the significant fluctuation and its impact on your overall savings pattern.

--