# Simple Random Sample and Non-response Bias

#### mysterdusk

##### New Member
Good evening,

I've the following question,

In a survey, a simple random sample of 1,000 households was drawn to determine the distribution of household size in a city. Interviewers were required to visit the house to carry out the survey. Interviewers could only reach 653 households in the sample. The rest of households was not reachable due to no one was at home to answer the house visit. That is, there was a non-response rate of 34.7%.

The researcher involved in this survey decided to draw a second batch of simple random sample of 1,000 households, and used the first 347 completed interviews in the second batch to bring up the sample up to its original planned size of 1,000 households. The survey result showed a total of 4,087 people in these 1,000 households, and the average household size in the city was estimated to be about 4.1 persons.

Choose one of the following options which is the most appropriate.

(a) The above described method of drawing sample would be a good way to fix the problem of non-response bias.

(b) The 347 households used in the second batch would be quite different from those unreachable 347 households in the first batch.

(c) The 653 households from the first batch would be a good representative of the population.

(d) The final sample of 1,000 household used in the survey can be considered as a random sample, as both batches of 1,000 households were drawn by simple random sampling method.

(e) The estimation of 4.1 persons per household in the city is likely to be about right.

My thoughts so far:

a) It is not a good way to solve the problem of non-response bias as households of a smaller size would have a higher chance to having nobody at home and thus, be underrepresented.

b) seems pretty possible as an answer.

c) Not true as the issue of non-response bias is left unaddressed.

d) also seems possible as an answer.

e) False due to reasons stated in a) and c).

For now I'm caught between b) and d), though I'm leaning more toward b).