Censored data & Truncated data
Censored data & Truncated data
Censoring occurs when an observation or a measurement is outside the range and people don’ t know the certain value. The value is always above or below the range that people set. However, truncated data means that because of the limits, such as time, or space, people lose some data. Truncation is to cut off the data.
In other words, we have collected and use the data, but the data is not in the range we have. It is called censored data. We don’t use the data because there are some limits, we don’t need to use and analyze the data, which is called truncated data.
Examples:
If I want to analyze the relationship between the sleeping hours per day and life span, I set the range of sleeping hours from 5 hours to 10 hours. If a person sleeps 12 hours a day, I will consider it is censored data because 12 hours’ data is not in the range. I will put this data into “over 10 hours”. Also, If a person only sleeps 3 hours, I will put this data into “below 5 hours”, and it is also censored data.
If the US want to import eggs from China, the US has the strict range of eggs’ size. The egg which is below the minimum size won’t be selected. That is truncated data.
Altman‘s Z Score
It is created by Edward Altman in 1968. Altman’s Z Score is a measurement to value the financial health of a company. And it can forecast the probability of the bankruptcy within 2 years.
Using the statistical method and a large amount of the cases of bankruptcies, Edward Altman created Z- Score formula:
Z= 1.2T1 + 1.4T2 + 3.3T3 + 0.6T4 + 0.999T5
In this formula, there are five business ratios:
T1 = Working Capital / Total Assets. This ratio measures the company’s liquidity and the features of the company’s scale. If working capital is less, it always means that the company is lack of turnover capital.
T2 = Retained Earnings / Total Assets. It reports the earning power of the company. The more retained earnings, the more power of paying dividends
Censoring occurs when an observation or a measurement is outside the range and people don’ t know the certain value. The value is always above or below the range that people set. However, truncated data means that because of the limits, such as time, or space, people lose some data. Truncation is to cut off the data.
In other words, we have collected and use the data, but the data is not in the range we have. It is called censored data. We don’t use the data because there are some limits, we don’t need to use and analyze the data, which is called truncated data.
Examples:
If I want to analyze the relationship between the sleeping hours per day and life span, I set the range of sleeping hours from 5 hours to 10 hours. If a person sleeps 12 hours a day, I will consider it is censored data because 12 hours’ data is not in the range. I will put this data into “over 10 hours”. Also, If a person only sleeps 3 hours, I will put this data into “below 5 hours”, and it is also censored data.
If the US want to import eggs from China, the US has the strict range of eggs’ size. The egg which is below the minimum size won’t be selected. That is truncated data.
Altman‘s Z Score
It is created by Edward Altman in 1968. Altman’s Z Score is a measurement to value the financial health of a company. And it can forecast the probability of the bankruptcy within 2 years.
Using the statistical method and a large amount of the cases of bankruptcies, Edward Altman created Z- Score formula:
Z= 1.2T1 + 1.4T2 + 3.3T3 + 0.6T4 + 0.999T5
In this formula, there are five business ratios:
T1 = Working Capital / Total Assets. This ratio measures the company’s liquidity and the features of the company’s scale. If working capital is less, it always means that the company is lack of turnover capital.
T2 = Retained Earnings / Total Assets. It reports the earning power of the company. The more retained earnings, the more power of paying dividends