If we have to understand how the people we are interested in, think on a specific topic, we need to conduct a survey. They are brief interviews with individuals on a specific topic. The way the researcher writes his questions so that different types of information can be collected, can be classified into various ways.
The different kinds of arrangement of questions for a survey would be:
2. Multiple Choice
3. Rank Order Scaling
4. Semantic differential scale
5. Stapel Scales
6. Open ended questions
1. Dichotomous: These are questions in a scale which give the respondents only two possible choices. Mostly here the options vary from yes or no. These kinds of scales are easy on scoring because of only two options and do the task of being good screener. However, some of the negatives of this kind of a scale are that they provide very little useable data because of limited information. The questions also offer only blunt specifics and nothing more.
2. Multiple Choice: In these kinds of questions, the respondent gets three or more choices that are mutually exclusive. The options here do not overlap. They are of course simple to understand and at the same time offer good choices to the researcher to draw more precise conclusions. However, its statistical application is limited and at the same time, it offers very limited gradation.
3. Rank Order: In this type of a setup, the respondent is expected to arrange the given options on the basis of their relative stature. The comparison of the alternatives can be done here. This type of questionnaire has the quality of being very specific and allows comparative statistics. However, sometimes here the list can get very large or specific.
4. Semantic Differential scale: this type of a scale uses a bipolar continuum by using two opposing adjectives. Some key features of this kind of a scale are that it gives specific valued information and also offers a neutral category but sometimes the neutral responses get too many and it is also required to score each item separately.
5. Stapel and Open ended Scales: They have their own characteristics but because they bring in a lot of subjectivity in the answers they become less relevant statistically as it is difficult to draw categories r grade such responses. Sometimes they become very vague as well.