Collection of Data

(i) Which of the following is the most important when you buy a new dress?
(a) Price (b) Fabric/Quality (c) Fashion/Design (d) Brand

(ii) How often do you use computers?
(a) Daily (b) Once a week (c) Once a month (d) Rarely/Never

(iii) Which of the newspapers do you read regularly?
(a) The Times of India (b) The Hindu (c) Hindustan Times (d) Other

(iv) Rise in the price of petrol is justified.
(a) Strongly Agree (b) Agree (c) Neutral (d) Disagree (e) Strongly Disagree

(v) What is the monthly income of your family?
(a) Less than ₹20,000 (b) ₹20,000 – ₹50,000 (c) ₹50,000 – ₹1,00,000 (d) More than ₹1,00,000

• Do you own a smartphone? (Yes/No)
• Is your school co-educational? (Yes/No)
• Do you have a library card? (Yes/No)
• Are you above 18 years of age? (Yes/No)
• Do you read a newspaper daily? (Yes/No)

Questionnaire on Vegetable Atta Noodles
Name: _________ Age: ____

1. Do you eat noodles? [ ] Yes [ ] No
2. Have you tasted Vegetable Atta Noodles? [ ] Yes [ ] No
3. If yes, how did you find the taste? [ ] Excellent [ ] Good [ ] Average [ ] Poor
4. Do you prefer Atta Noodles over Maida Noodles? [ ] Yes [ ] No
5. Will you buy it again? [ ] Yes [ ] No

• (i) There are many sources of data.
  True. (Primary and Secondary sources).
• (ii) Telephone survey is the most suitable method of collecting data, when the population is literate and spread over a large area.
  False. (Mailed questionnaires are generally considered most suitable for literate populations over large areas due to lower cost, though telephone surveys are also used).
• (iii) Data collected by investigator is called the secondary data.
  False. (Data collected by the investigator himself/herself is Primary Data).
• (iv) There is a certain bias involved in the non-random selection of samples.
  True. (Judgment/Quota sampling depends on the personal prejudice of the investigator).
• (v) Non-sampling errors can be minimised by taking large samples.
  False. (Sampling errors decrease with sample size, but Non-sampling errors [errors in data collection/processing] may actually increase as the volume of work increases).

• (i) How far do you live from the closest market?
  Problem: Ambiguous. The term “far” is subjective. Better: “What is the distance in kilometers from your house to the market?”
• (ii) If plastic bags are only 5 per cent of our garbage, should it be banned?
  Problem: Leading Question. It provides information to bias the answer towards “No”.
• (iii) Wouldn’t you be opposed to increase in price of petrol?
  Problem: Leading/Double Negative. It pressures the respondent to say “Yes”.
• (iv) Do you agree with the use of chemical fertilisers?
  Problem: Vague. It is not specific about “use” (usage or policy).
• (v) Do you use fertilisers in your fields?
  Problem: This is a decent two-way question, but lacks specificity on the type of fertilizer (Chemical vs Organic).
• (vi) What is the yield per hectare in your field?
  Problem: Calculation required. Many farmers may know total yield but not per-hectare yield, leading to inaccurate data.

• Population: 200 farms (The aggregate of all units).
• Sample Size: 50 farms (The subset selected for study).

• Population: (1) All students in a school. (2) All cars produced in a factory in 2023.
• Sample: (1) 10% of students selected for a health checkup. (2) 50 cars selected for quality testing.
• Variable: (1) Height of students. (2) Mileage of cars.

It depends on the context, but Sample Method is generally considered to give “better” results in a practical sense because:

• Cost & Time: It is economical and faster.
• Operational Efficiency: Fewer items allow for more intensive study and better-trained enumerators, reducing non-sampling errors.
• However, if 100% accuracy is required (e.g., National Population Census), the Census method is better.

(b) Non-Sampling Error is more serious.
Reason: Sampling errors can be estimated and minimized by increasing the sample size. Non-sampling errors (errors in data collection, recording, or bias) are difficult to detect and cannot be eliminated simply by taking a larger sample; in fact, they may increase with sample size.

*(Note: The question likely implies “Sample Surveys vs Census Surveys”).*
Yes, Sample surveys often provide better results because:

• Reduced Non-Sampling Errors: Fewer units mean enumerators can be better trained and supervised.
• Feasibility: In cases where testing destroys the unit (e.g., testing bullets or chalk), sampling is the only option.
• Economy: It saves vast amounts of time and money.

The number of possible samples is calculated using the combination formula $^nC_r$, where $n=10$ and $r=3$.
$$ \text{Samples} = ^{10}C_3 = \frac{10 \times 9 \times 8}{3 \times 2 \times 1} = 120 $$
There are 120 possible samples.

Procedure:

1. Write the names or roll numbers of all 10 students on separate, identical slips of paper.
2. Fold the slips so names are not visible and place them in a bowl.
3. Shuffle the slips thoroughly to ensure randomization.
4. Ask a neutral person to pick out 3 slips one by one without looking.
5. The students whose names appear on the slips form the sample.

Yes, provided that:

• The slips are identical in size, color, and shape.
• The mixing is thorough.
• The selection is done blindly without any bias.

If these conditions are met, every student has an equal chance of being selected, ensuring a random sample.

Procedure:

1. Assign a serial number from 01 to 10 (or 0 to 9) to every student in the class.
2. Take a standard Random Number Table (e.g., Tippett’s table).
3. Blindly point to a number on the table to start.
4. Read the numbers in a sequence (row-wise or column-wise).
5. Select the first three numbers that fall within the assigned range (01-10). If a number repeats or is outside the range, skip it.
6. The students corresponding to these numbers form the sample.