Collection of Primary Data – Methods & Precautions
Primary Data = Fresh Data! Like cooking a meal from scratch instead of ordering takeout. 🍳
What is Primary Data?
[!NOTE] Definition: Data collected by the investigator for the first time for a specific purpose. It is original in nature.
Also called: First-hand data, Original data.
Example:
- A company conducting a survey to know customer feedback.
- Census of India collecting population data.
Methods of Collecting Primary Data 📝
There are 5 main methods. Let's compare them!
1. Direct Personal Investigation (Interview) 🎤
Method: The investigator meets the respondents face-to-face and asks questions.
Suitability:
- When intensive study is needed.
- When area is small.
- When accuracy is critical.
| Merits ✅ | Demerits ❌ |
|---|---|
| High Accuracy: Investigator can verify facts. | Expensive: Travel cost is high. |
| Response Rate: People rarely refuse face-to-face. | Time Consuming: Takes long time. |
| Flexibility: Can explain/clarify questions. | Bias: Investigator's bias may affect results. |
Example: Police interrogation, Census enumeration.
2. Indirect Oral Investigation 🗣️
Method: Information is collected not from the person concerned, but from third parties (witnesses) who know the facts.
Suitability:
- When the person is reluctant (e.g., drug addicts, criminals).
- When the area is large.
- When direct contact is impossible (e.g., fire accident).
| Merits ✅ | Demerits ❌ |
|---|---|
| Economical: Saves time and money. | Less Accurate: Third-party info may be wrong. |
| Wide Coverage: Can cover large area. | Bias: Witnesses may be biased. |
| Expert Opinion: Can get expert views. | Reliability: Depends on witness honesty. |
Example: Police asking neighbors about a theft, Commission of Inquiry.
3. Information from Correspondents 📰
Method: Local agents (correspondents) are appointed in different areas to send regular reports.
Suitability:
- When regular/continuous information is needed.
- For newspapers, magazines, government agencies.
| Merits ✅ | Demerits ❌ |
|---|---|
| Very Cheap: Low cost. | Low Accuracy: Rough estimates only. |
| Continuous Data: Regular updates. | Uniformity: Different agents use different methods. |
| Wide Coverage: Covers entire country. | Delay: Reports may come late. |
Example: News channels getting reports from local reporters, Crop price reporting.
4. Mailed Questionnaire Method ✉️
Method: A list of questions (Questionnaire) is sent by post or email to respondents. They fill it and send it back.
Suitability:
- When area is very large (National survey).
- When respondents are educated.
| Merits ✅ | Demerits ❌ |
|---|---|
| Cheapest Method: No travel needed. | Low Response Rate: People ignore mails. |
| Wide Coverage: Can reach anywhere. | Inflexible: Cannot clarify doubts. |
| Originality: No investigator bias. | Illiterates: Cannot use for uneducated people. |
Example: Customer satisfaction surveys via email.
5. Schedules Sent through Enumerators 📋
Method: Enumerators (staff) go to respondents with the questionnaire (Schedule), ask questions, and fill it themselves.
Suitability:
- When respondents are illiterate.
- When area is large but funds are available.
- For government surveys.
| Merits ✅ | Demerits ❌ |
|---|---|
| High Response: Enumerator ensures filling. | Most Expensive: Salary + Travel. |
| Suitable for Illiterates: Enumerator writes. | Time Consuming: Takes years. |
| Accuracy: Trained staff fill data. | Bias: Enumerator bias possible. |
Example: Population Census of India, NSSO Surveys.
Comparison of Methods 📊
| Method | Cost | Time | Accuracy | Coverage | Suitability |
|---|---|---|---|---|---|
| Direct Personal | High | High | Highest | Small | Intensive study |
| Indirect Oral | Low | Low | Low | Large | Complex issues |
| Correspondents | Very Low | Low | Low | Very Large | Regular info |
| Mailed Questionnaire | Low | Medium | Medium | Large | Educated people |
| Enumerators | Very High | High | High | Large | Illiterates/Govt |
Essentials of a Good Questionnaire ✅
If you use the Questionnaire method, the form must be perfect!
- Cover Letter: Explain purpose politely.
- Number of Questions: Minimum possible (15-20).
- Simple Language: No technical jargon.
- No Personal Questions: Avoid income, age (unless necessary).
- Logical Sequence: Easy → Difficult questions.
- Types of Questions:
- Simple Alternative: Yes/No
- Multiple Choice (MCQ): A/B/C/D
- Open Ended: "What is your opinion?" (Avoid if possible)
Example of Bad Question: "Don't you think the government is corrupt?" (Leading/Biased) Example of Good Question: "How would you rate government performance? (Excellent/Good/Poor)"
Summary
- Primary Data = Original, first-hand.
- 5 Methods: Direct, Indirect, Correspondents, Mailed, Enumerators.
- Best for Accuracy: Direct Personal.
- Best for Large Area: Mailed / Correspondents.
- Best for Illiterates: Enumerators.
- Questionnaire: Should be short, simple, and logical.
Exam Tip: Remember the difference between Questionnaire (filled by respondent) and Schedule (filled by enumerator)! 📝
Test Your Knowledge
Question 1 of 5
1. Data collected for the first time is called: