This is more of a theoretical issue, my boss and I are having a disagreement about how to design a certain questionnaire.

There are 76 questions, so it is rather long. Normally we go question by question, my boss under the belief that having to scroll down to answer questions will annoy people.

I think having to constantly click to get to the next question is annoying.

What do people think? More pages or longer pages?

I'm inclined to side with your boss on this one I'm afraid..

The thing you have to remember is, people are IDIOTS! And there is bound to be at least one clown out there who can't figure out the scrolling mechanism on his/her mouse.

Another point in your bosses favour - the percentage complete bar is relative to what page you are on, not what question you are on.

But hey, that's only my opinion. Everyone else may disagree.

Mike

Clicking "next" 75 times also sounds quite annoying. I'd suggest to group questions to have 1-4 questions on one page wherever that makes sense.

I'd go with your boss on this one, as well.

I always hate it, when I have to scroll through long questionnaires. I'd rather prefer to click a few more times. We always try to do the question-per-question way. Sometimes there are shorter questions, which could be joined together on one page (unfortunately this is only possible by going the group-by-group way).

Page per page or group per group also is a lot better when there is a certain skip rules or conditions in the questionnaire.

If you have a look around, most professional market research companies avoid online questionnaires with a lot of questions at once.

What some tools offer to avoid a lot of clicking is for questions with single answers to forward directly to the next page when one answer was picked (which obviously has some disadvantages aswell).

I'm with Mazi on this one. I typically place as many questions on a page as will fit into an 800x600 screen. That way almost nobody will need to scroll and the number of clicks is drastically reduced.

Having said that, one needs to consider question context (does it make sense to group the questions together) and, as holch mentions, skip patterns (it can be annoying to have questions pop into and out of existance as one answers the survey).

And besides, you should never agree with your boss

> And besides, you should never agree with your boss

Shouldn't this be "never disagree"?

It probabla depends on your boss