Question

In each of the following areas, what're two examples of how data accuracy can be compromised by inappropriate choices of the researcher;

a. Survey questionnaire design

b. Sampling plan for a survey

c. Data collection and recording

Answer 1

a) One error of survey questionnaire design can be when the questions are too complicated or vague such that the interviewers might not understand it and even the people recording might understand something else. This leads to incorrect responses and results.

Another type of error for this type is when the questions are overlapping having common answers and unclear where instead of being sure of the answers there is a chance of choosing among the two. This will lead to confusion among the people who answer.

b) A sampling plan error can be a population specification error, when the researcher does not understand who they should survey and selects the wrong population. For example a new product is supposed to be launched so before that the company conducts a survey to find out potential customers. They can survey all the people who consume the product instead of the people who actually purchase and use the product. So this might give more positive results to the company as the people who consume the product would be higher than people who buy it.

A sampling frame error could also be Missing a particular group of the population who are actually a part of the population. While missing surveying them changes the results as their responses are also Population of interest.

c) There might be errors in recording the responses by the person who interviews (say he notes different answers than what has been said to him), this changes the results and is called interviewer error.

Another type of error of this is selection of a non Random sample. This leads to biases as the responses of the group might be affected by each other and it's not independent or accurate. This is called a data selection error.