Article Index

You will not like what comes next. In many cases, a questionnaire asks for information that a service provider could use to identify you. Or it asks for information useful for social engineering. Often, of course, genuine market research, opinion polling, or substantial scientific work is taking place. In other cases, however, it may involve abuse, spearphishing, or some other dirty trick. The best case is that there is no dirty trick: the questionnaire only costs you time, and you receive something in return. Or at least a promise, a little illusion. Sometimes the questionnaire costs you aggravation as well as time. We have seen examples of that too. And then, of course, there are the truly dirty tricks.

One questionnaire does not reveal too much dangerous information about you. But what about five? And what if they can see your Facebook profile too? Questionnaires are often used for more than asking a community's opinion. The situation above is interesting because the same kind of questionnaire can paint a “standardized” picture of individual characters' backgrounds. Or at least of what their players think about their characters. If I had to show a sample questionnaire now and then explain what was arranged how, and why, I would start with a questionnaire that was bad in many respects. Not just bad, but downright awful, immediately raising the question: why? Because I would deliberately restrict the possible answers, deliberately manipulate. The questionnaire's answers would distribute points according to the character creation rules, producing a character who fit the adventure. To do that, however, I would have to exclude answers that really did not fit the adventure.

And this would already be crucial when choosing the right platform. The program must be able to calculate: that is a basic requirement. There must be read-only fields. There must also be...

Imagine a series of questionnaires. They appear to concern transport habits. The first few batches of questions concern your present transport habits. They ask about many service providers, then make an interesting little shift. They try to compare these with your childhood transport information. Let us say the respondents live in Budapest. A district alone is not much information. But the kind of vehicle you used to reach a metro station then, and the kind you use now, is already important information. Approximately how many stops you traveled, how long you traveled, whether you did, what proportion of the route you covered: these still examine transport information. So do questions measuring your journey to school instead of to the metro. But the first set tells them with fairly good accuracy where you lived, while the second helps them work out where to look for the school. They may know the year from demographic information too.

How long does it take, searching Facebook and asking former school staff and other pupils a few questions, to find out “what was the first-grade homeroom teacher called in year X?” Perhaps for several first-grade classes. This can easily supply the answers to two secret questions. If they ask you later, perhaps on another platform, questions about banking habits and many other things can follow. Naturally, in each case they know your IP address and numerous details needed to identify your computer. And they look for a nearby Wi-Fi network they can easily join, so the attempt does not even come from an unfamiliar location. The browser's user agent and many other identifiers may be familiar, since these can be copied. 

Connections between questions in a questionnaire, or connections between several linked questionnaires, can both be used to extract information you do not want to disclose. Indeed, with a little luck, filling out a questionnaire can reveal something you otherwise would not tell them simply because you do not know the answer for certain yourself. The problem is the questionnaire's length. And many people do not welcome having their private lives intruded upon and analyzed. This is another reason they reject questionnaires. Of course, that is only true up to a point. There is another mechanism.

The simplest situation is when the reward for completing the questionnaire is neither tangible, nor a promise, nor some result, but the knowledge that you did not spoil anything good; you even helped by spending those five minutes on it. And anyone who says the whole thing is bad and was not even worth that can go screw themselves. After all, they have spoiled the good feeling that even completing the most awful questionnaire gave you. The next situation is when you receive a promise or an illusion in exchange for the questionnaire: this will improve something you care about too. For D&D's fifth edition, every playtest packet came with a questionnaire, and we filled them out; we even provided feedback on the forum. But many of us do not play the finished game. The enthusiasm generated on the basis described here can very easily turn into disappointment: all it takes is for them to call something final that you intensely dislike, when they did not even ask about the most important arguments and considerations.

Of course, someone lectured me that the thing I complained about no longer worked that way in the final PHB. They simply failed to understand why this did not interest me. Yet it is very simple: supplements will not have a test version, something that sweeps away the supposedly final solution. I would like to believe the questionnaire and testing served a purpose. But it is damned easy to feel disappointed when the questioner fails to notice one thing: Charm Person was not the only iconic thing in D&D; many races also had various traits and limitations that justified those traits. When these slowly disappear, and the explanation amounts to “well, there are many worlds, so we cannot include the limitation,” I have reason to be disappointed. Because, excuse me, this is someone saying that a longsword costs the same in every city in every world, that abilities are universal too, that the same spells were discovered everywhere—and yet the iconic things that matter to me are precisely what is missing.

And here I should point out that the D&D questionnaire was merely a simple example. With a local café too, it is natural for a questionnaire's main promise to be that the product will become better for us, that we will benefit. Of course, there too, one person wants a vegan café with strongly feminist politics, another wants Hooters, so someone will be disappointed. But a questionnaire put out on the premises can certainly be used for profiling too. Unlike the D&D questionnaire, however, this may often be on paper; asking for a few minutes that way is completely different from an online questionnaire linked in local Facebook groups. It is no accident that offline questionnaires often come with an interviewer. 

Most people say you cannot ask everything. They are right. You cannot, should not, and above all need not ask everything. People will give you five minutes out of goodwill. But fifteen minutes usually needs a promise behind it, and half an hour or an hour needs a substantial reward. Yet even a moderately amateur practitioner will say: a question in a questionnaire can easily control which questions appear later, which ones we must answer. If, instead of first asking which spell is iconic and making me check Charm Person, they ask what features of (A)D&D made me play it, offering the world, storyline, spells, special races, and so on, they will probably not subsequently ask me how iconic Charm Person was. But they will learn about the time it took ten to fifteen adventures to get around a racial limitation because our dwarf had made a sacred vow to become a wizard if the god of magic healed his friend; or how iconic it was when the elf girl, thanks to her wild wood-elf heritage, could lift more than the heroic knight could move, though her strength did not show. And how much, how very much we laughed when someone wanted to help her. Her weapons did not even take advantage of her strength. But she made artistic weapons. 

Asking only a few questions on the grounds that “there are many possible answers, it took a lot of typing” is utterly embarrassing. We do know where that mistake was made, do we not? A good questionnaire needs branches and a relatively large number of questions. Enough to connect answers that say little individually but become revealing together. In offline questionnaires, the interviewer often handled these branches. Their most important characteristic, of course, was that they did not influence the respondent. At least in theory. In many cases, however, if someone assumes the interviewer has very different opinions from theirs, the wish to meet the interviewer's expectations may still arise. And that brings us to the next part of the subject.