91Ó°ÊÓ

The median is a measure of central tendency that represents the middle value in a dataset. It is particularly useful when dealing with skewed distributions or outliers. To find the median, you first need to arrange the data in ascending order. Once sorted, the median is the value that separates the higher half from the lower half of the dataset.
In our exercise, we had eight lifetimes available for calculation:

• 170
• 290
• 350
• 480
• 570
• 790
• 860
• 920

To calculate the median, we identify the fourth and fifth values because with eight numbers, the median is the average of these two middle numbers.The formula for calculating the median here is \(\frac{480 + 570}{2} = 525 \)
As you can see, despite the presence of censored data (the '1000+') in the dataset, the median can still be calculated without being skewed by these values. That makes it a robust measure of central tendency in such scenarios.

Censored data occurs when some data points in your experiment or study do not have a precise value, often having a limit or threshold instead. In this exercise, the data points marked as '1000+' indicate that after 1000 hours, the light bulbs had not failed yet.
Censored data can pose challenges in statistical analysis since these values are not exact and could be potentially large values. However, this makes measures like the mean more difficult to use directly since the true values are unknown. For instance, while we know the bulbs surpassed 1000 hours, they could have lasted significantly longer.
Handling censored data requires special statistical methods and careful thought about how best to interpret results. The median is a measure that remains unaffected by right-censored data points, which is why it is often preferred in such cases.

The mean, also known as the average, is calculated by adding all data points and dividing by the number of data points. While it is a very common measure of central tendency, the presence of censored data can distort its accuracy.
In the given exercise, the inability to know the exact values of '1000+' bulbs affects the mean significantly. If these bulbs had lasted much longer, their inclusion in mean calculations could potentially raise the average value.
Therefore, the mean becomes a less reliable measure when dealing with censored data. The uncertainty introduced by the '1000+' values means the mean does not necessarily reflect the true average lifetime of all bulbs. This is why practitioners turn to means specifically designed for censoring issues or alternative statistics, like the median, which remains unaffected by such outliers.