The Psychology of Quality and More

A Toolbook for Quality Improvement and Problem Solving (contents)

Surveys: How to understand it

The Quality Toolbook > Surveys > How to understand it

When to use it | How to understand it | Example | How to do it | Practical variations

<-- Previous | Next -->

How to understand it

In many situations, information required about a problem is held by a broad group of people, such as female customers or company engineers. Simply asking them about the problem without any thought of how their replies will be used is likely to lead to wrong conclusions and bad decisions. Good decisions require careful data collection and analysis.

'Surveys' is a general term for a number of methods of collecting data from people, which includes questioning them individually and collectively, in person, over the telephone or on paper. The decision of what type of survey to use is driven by such factors as the accessibility of the people and the response required. For example, in-depth personal interviews can give much useful information, but may be impractical when there are many people to ask.

There are two main approaches to surveys, interviews and questionnaires, as contrasted in Table 1. Interviews involve one or more interviewers asking questions of one or more respondents, whilst a questionnaire contains written questions and answers. The most noticeable difference is that with a questionnaire, the person asking the questions and the respondent answering them may never meet.

Table 1. Interviews vs. Questionnaires

A problem with asking people questions is that not only will different people answer differently, but changes in the wording, sequence and environment of questions are likely to result in different responses from the same person. The design of a questionnaire and the execution of an interview must therefore be done with great care.

The two main types of questions that can be asked in surveys, closed questions and open questions are contrasted in Table 2.

Where there is a large population to question, it is often impractical to survey everyone and care must be taken to select a random sample. As with other tools that use sampling, a non-random selection can give biased results.

The opinion that people hold is often heavily influenced by their professional or social peer group. If the results of a survey of a wide group are just summed or averaged, then these differences may be lost. It is thus important to recognize and differentiate between clustering of results, either by questions asked in the survey, or, if different questions need to be asked, by separately surveying recognized groups.

Surveys often require a significant amount of effort in preparation, execution and analysis. However, this is often the best method of acquiring the information needed to make reliable decisions.

Table 2. Open vs. closed questions

<-- Previous | Next -->