Hi Terry, it goes to show that there are no right answers to these things. One piece of conventional wisdom was to put them at the end. There were several reasons for this. One was sometimes people are hesitant to give answers to demographic questions, so the idea was to get them invested in the survey. Another was that they might get annoyed by being asked such questions and if you lead with those you might lose them as respondents. I think that if there are more than a couple there is definitely a point to the second issue. Another thing to think about is that there is considerable research on the ordering of questions (although I don't know how much has probed this), and it does affect how people answer. It may make a difference in how people respond if they know that their characteristics, let's say age, race, gender, are being taken into account. I don't have any specific article to recommend, but sometimes there are interesting articles about surveys in the Public Opinion Quarterly. Bob

Hi Terry, I was also trained to put them right a the beginning. Over the years in my survey research experience I learned that I can also use simple ones like gender, age and perhaps where you live (like rural urban etc) and ask other questions like household information, family income, insurance, other somewhat sensitive questions with in the context throughout the survey so they do not feel like they are answering too many personal questions. This works for me when my subjects are from diverse populations. Hope this helps.

Since some participants in research projects sometimes seem to only answer half the questionnaire/or not the end - it seems important to add the most important questions such as gender, age etc, in the beginning so at least this information is captured. At least that have been the rationale in some research projects I have been a part of.

However, some argue that since demographics are easier to answer, they should be in the end of the survey when the participants are fed up with thinking about what to answer - and thereby save the beginning for more cognitively difficult questions.

However, I think I would prefer to have age, gender, living conditions on the first page to start them off in a easy manner, and have household economy/income/similar on the second or third page. So, agreeing with @Susie Ramisetty-Mikler in not grouping the personal questions.

Hi, just to mix it up. A lot will depend on the type of survey. I am primariy concerned with work-life research and tend to put demographic questions relating to work such as education, tenure, worktime etc, in the work section and demographic questions relating to family e.g. age, gender, children, civil status in the section relating to family/life issues.

If I am interested in well-being or health issues I usually try to put these questions near the end of the survey. I prefer this method but I don't think there is a strict, scientific method to help you make this decision. My suggestion would be to study the designs perferred by others in your area of research and do what they do.

This is an interesting issue. I was trained to put them at the beginning. At least you can gauge age of consent and sex differences for participants who have petered out and not completed the questionnaire. I came across this topic when I was looking for literature on best practices. It seems there is no one correct answer. Read the opinions posted here has been valuable.

I think it should come at the end of the questionnaire after the main measures.

What we think is meaningless, right? I train my students to design two versions of the survey to test for order effects. Good topic though.