Loss of information
Topic: Economic Data: Census, NSS, Surveys and Statistical Tools · NCERT: Class 11, Ch 3 "Organisation of Data"
Meaning
Loss of information is the main weakness of grouped data. Once values are put into classes, we no longer use the actual values. Every observation in a class is treated as equal to the class mark, the midpoint: (upper limit + lower limit) ÷ 2. Grouping makes a large mass of data short and easy to read. The price is that some detail is lost. So averages worked out from grouped data are only close estimates of the true values.
Example
Six values, 20, 22, 25, 25, 25 and 28, fall in the class 20-30. After grouping, all six are treated as 25, the class mark. Their true average is 145 ÷ 6 ≈ 24.17, but the grouped data give 25.
Don't confuse with
- Sampling error: this comes from studying only part of the population. Loss of information comes from grouping, and it happens even when every unit has been counted.
Related concepts
- Raw data
- Classification
- Chronological classification
- Spatial classification
- Qualitative classification
- Quantitative classification
- Attribute
- Continuous variable
- Discrete variable
- Frequency