1. Errors in Sampling

When we collect data from a sample instead of the whole population, errors can occur. These errors are broadly classified into two types:

1. Sampling Errors
2. Non-Sampling Errors

2. Sampling Error

Definition: Sampling error is the difference between the sample estimate and the true population value that occurs because only a sample, not the whole population, is surveyed.

Characteristics:

• Occurs naturally due to selecting a subset
• Can be reduced by increasing sample size
• Cannot be completely eliminated

Example:

• True population mean of student heights = 165 cm
• Sample mean = 163 cm
• Sampling error = (165 - 163 = 2) cm

Key Point:

• Sampling error is random and statistically measurable.
• Often expressed using standard error.

3. Non-Sampling Error

Definition: Non-sampling error is the error that occurs not due to the sample size, but because of other factors in data collection, processing, or measurement.

Sources:

1. Measurement errors – Faulty instruments, wrong recording, biased questions.
2. Non-response errors – Some selected units do not respond.
3. Processing errors – Mistakes in coding, tabulation, or data entry.
4. Selection bias – Sample does not represent population correctly.

Characteristics:

• Can occur even if whole population is surveyed
• Often systematic, leading to bias
• Can be reduced with careful planning and supervision

Example:

• A questionnaire asks: “Do you always recycle?” Some people may over-report due to social desirability bias.
• Some selected households could be unreachable, causing non-response error.

4. Comparison Table

5. Key Points to Remember

1. Sampling errors are unavoidable but estimable.
2. Non-sampling errors can be larger and more serious than sampling errors.
3. Good survey design, pre-testing, and training reduce non-sampling errors.
4. Always distinguish between random errors (sampling) and systematic errors (non-sampling).