Open-ended and closed-ended questions. Open-ended response allows respondents to write down whatever comes to mind and to express a response how they like. Closed-ended questions allow only a limited range of answers for the respondent.

Ranking questions generally ask respondents to rank different qualities from most important to least important entailing the inclusion of all values in a response format, whereas rating questions generally require that respondents select only one or a limited number of responses to indicate a state.

Coding is the systemic scoring of survey questions and answer categories. Coding is important because the scoring system or codes produce the statistic from the variable and also, depending on the number and type of answer categories, they can determine the level of measurement of a variable.

The level of measurement of a variable is determined by the question type and answer category. For example, the difference between a Likert scale and the level of income.

Because they can cause confusion for respondents leading to poor data quality due to inaccurate answers.

The problem is that the respondent and researcher cannot be clear about which category or reference point in the question is meant to be the main category of the question. These questions can be improved by separating out the two or three categories of interest in the question.

Pilot testing is a program of asking a limited number of respondents to complete a survey and report any difficulties they had experienced while doing so. It is important for survey design because this process can identify problems with question construction and progress through the survey.