📉If two different samples of the same size from a population give very different results, what might be the reason?

📉If two different samples of the same size from a population give very different results, what might be the reason?

Answer: This situation occurs when sampling variability or non-sampling errors are high.

Possible Reasons:

1. High Sampling Variability:

   • Population is highly heterogeneous.

   • Sample size nnn is too small.

   • Sampling method is inefficient. Mathematically, sampling variance

     
     

2. Improper Sampling Method:

   • Non-random or biased selection.

   • Poor stratification or incomplete sampling frame.

3. Non-Sampling Errors:

   • Measurement mistakes, wrong recording, or interviewer bias.

   • Non-response leading to unbalanced data.

4. Outliers or Extreme Values:

   • Presence of few high or low values may distort sample mean.

Remedies:

• Increase sample size (reduces variance).

• Use stratified or systematic sampling.

• Improve questionnaire quality and field control.

  
  

Cross-Question: 👉 If variability remains even after increasing sample size, what should be checked?

→ Non-sampling errors (bias, frame errors, or recording mistakes) must be investigated.

Context Example: When NSS collects consumption data, large inter-sample variation may occur due to regional heterogeneity; hence, stratified design and larger sample size are used to stabilize estimates.