Tuesday, April 6, 2010

The best questions about stories - more

This is a follow-up post to "The best questions about stories" with a few more thoughts. I said in that post that I've observed four dominant factors that make questions about stories work. Since then I thought of two more factors (and also at the end I have a little more advice on reducing the number of questions).

5. Mind your messages.

It is so very easy to slip subtle messages into questions about stories. We all betray ourselves when we ask questions; it's human. Some messages, like "I am excited about this project" and "I hope you will find yourself heard here" and "I'm here to listen" and "I respect your viewpoint" are messages you want to communicate. But other messages can creep in, ones that damage the potential of the project. Those are worth finding and removing.

I can think of two types of message you can send in your questions about stories: directing messages (what to do) and characterizing messages (who you are, who we are).

Directing messages.  A directing message tells people which responses are acceptable and which are not. This is close to a "leading question" in direct questioning surveys. The most extreme directing message is what I call the "do you want to keep your job" question set, in which the correct answer to each question is painfully obvious. Projects with question sets like this are foregone conclusions and aren't worth finishing. But there are many lesser grades of directing questions that can slip under your radar.

Say you are asking patients for stories about their relationship with their doctors. Say you ask what the story shows about the doctor's committment to quality medical care. If all of the available answers are positive, ranging from "satisfactory" to "excellent" for example, you have communicated to your storytellers that you do not want them dishing any dirt on their doctors. "Quack" is not a possibility, nor is incompetent, dishonest, burned out, lost, or confused. The message is: you will say this, and you will not say that. The worst part about a message like this is that it will affect not only the answers to that question but every answer and every story people tell you after they see it.

Or, say you are looking at innovation among the employees of your organization. One of your questions has to do with what the main characters of stories do with their work time. It would be a project-killer to include "play" or "chatting" or "messing around" as possible answers, even though this is the sort of thing you are looking for, because nobody in their right mind would choose that. But if you put in an answer like "unconstructed exploration" or "collaborative brainstorming" you are likely to find the play you are looking for.

This is not to say you should not direct people when you ask questions about stories. Certainly you want to nudge people! You want them to reflect on their story, to consider the questions carefully, to recall the information you need, to explore their feelings. Keep those directions and lose the others. My suggestion is to have a range of people read the questions and report on their feelings about where they feel nudged. Are they nudged in the way you'd like them to be?

Characterizing messages. A characterizing message comes in where you make a statement about the worldview of your storytellers and, for some of them, your statement doesn't fit. To those people that constitutes a message, intended or not. The message is: this is not for you.

The classic example of this sort of characterization is in the puzzle that goes like this.
A man and his son are rushed to the hospital, both badly hurt. The surgeon on duty in the emergency room cries out, "I cannot operate on this boy! He is my son!" 
At this point so many people know this story that it's just a funny little test to see if there's anyone left who can't guess the answer. But think how the surgeon feels. And think how the surgeon would respond to questions that assume she is a man. What would she do? She might try very hard to prove she can still be considered worthy of speaking; she might lose enthusiasm and plod through the rest of the required task without giving due consideration to anything; she might angrily answer the questions vindictively; or she might simply walk away and exclude herself from the project. All of these responses are misfires for a narrative project.

A few more examples. Your list of possible answers to the question "How do you feel about this story" shows people what sorts of emotions you expect them to have. Say the list includes "hysterical" but not "calm" or "motivated" but not "desperate" or "happy" but not "sad": that communicates a message. I like to check for counter-balancing answers. If there is a "desperate" there should be a "reassured" or "content." If there is "to create" there should be "to destroy." Or, if there is a long list of possible problems, one answer should be "there wasn't a problem." Why? Because if you don't have that answer the message is "you are a person who is guaranteed to have problems."  

Also watch out for "of course everybody thinks this" or shared-value messages. These make a statement about the worldview of your storytellers by describing your worldview and making it clear that you assume they share your opinion. Again these can be so subtle that you don't notice them. Even the names you use for places and people can communicate presumably shared values. Listen to this sentence from Wikipedia on the name of one of the airports in Washington, D.C.:
The airport is commonly known as "National", "Washington National", "Reagan", and "Reagan National".
Which of these names you call that airport has something to do with your feelings about Ronald Reagan and his politics, and your age and history in the area. Every city has a few of these tellingly-named landmarks. Even the names of some whole cities, and countries, are indicative (Mumbai or Bombay? Burma or Myanmar?). Most people in official roles can also be referred to in several ways. Is a teacher different from an instructor? How about an educator? A trainer? What is the difference between a cop, a policeman, a police officer, a law enforcement agent, and an officer of the law? This page on Wikipedia lists an amazing 119 (at this count) slang terms for police officer. (Wow. Important?)

It is so easy not to notice characterizing messages that I suggest always having someone else look for them. Of course you can't write questions that encompass the worldview of every sentient being on the planet, but you can aim to get to the point where most of your storytellers believe the project is about them and not someone else.

6. Gather interpretations, not opinions.

When you ask questions about stories, you want people to direct their attention to the story and away from themselves. This creates that magical narrative displacement you need to get past defensive walls and gather authentic ground truths about the topic you are exploring. So make sure your questions keep people engaged in interpretation and don't allow them to wander over into opinion. For example, even if you are one hundred percent sure that each of your storytellers will be the protagonist in their own stories, don't ask a question like, "In this story, what did you need to solve a problem?" Doing that will remove them from the context of interpretation and place their attention back on themselves. Instead, ask "In this story what did the main person need to solve a problem?" Keep them in the story.

It's a bit tricky figuring out how to refer to whoever is in the story without saying "you" to people. Rarely can you actually use the word "protagonist" because most non-screenwriters will not understand what you mean. Instead it's better to talk about the main or central person or character. Character is more broadly understood than protagonist, but it can be a dangerous word to use with suspicious storytellers or sensitive topics, because it can be taken to imply that the story is a fiction or lie. Person, though not very exciting, is the least risky.

When you ask questions about the story's main character, make sure to ask who that is. I've seen a few question sets that asked about behaviors, beliefs and characteristics of the central person, but then forgot to ask who that person was. When this happens the answers get mixed together -- was it the teacher or the student who showed compassion? was it the sales agent or the customer who needed information? -- and lose their utility. I hate it when this happens, because people often reveal things about themselves as a story character that they are less willing or able to reveal directly. A simple question like "Who is the central person of this story?" (and then some choices based on the important societal/topical roles of people who might be in the story) can avoid this problem.

You might think that the question "How do you feel about this story?" presents an exception to the keep-their-eyes-on-the-story rule. But it doesn't. If you asked the question "What does your story mean?" or "What is your story about?" you would open the gates to the world of platitudes and close down the path to ground truth. Don't ask people why they told their story. If you do that, you might as well skip the story and go straight to "what do you think about this issue." Instead, ask people what the story says to them, what emotions it brings up. That's still interpretation, and that's still magic.

A few more tips on reducing the number of questions

In nearly every project in which I've helped people design questions, we end up with more questions than we can use. That's a good sign. It means you are thinking broadly and will pick up a wide range of useful information. If you don't have to trim your question number you haven't been thinking deeply or aspirationally enough. But trimming them down can be daunting. Every question wants to be asked; but if you ask them all, you ask none of them.

Here's a method that works. Draw a table on a large piece of paper, or use a spreadsheet. On the rows and the columns write your questions (twice). Give each question a number. Now compare the questions in pairs. For each pair ask yourself: if you could only have one of these two questions, which would you choose? Write the number of the winning question in that cell of the table. As you consider each pair, also think about their proximity in conceptual space. Are these two questions similar? If so, why do you have them both? Were you trying to get at something slightly different in each of them? Can they be merged without losing anything? Or should they be moved further apart? You might have to iterate parts of the table a few times if you are merging and splitting questions. (Gotta love those spreadsheets.)

After you finish your table, count up the number of "wins" for each question, and you'll have a ranking. If you can only use seven questions, choose the top seven on the list. Some questions require others, so you may not be able to simply cut off the list so easily, but in general this method can get you through the obstacle of staring at the questions not knowing which to pick. (I picked up this technique long ago from Richard Bolle's What Color is Your Parachute, and I'll bet a lot of people recognize it and have used it in other contexts.)

One thing I've noticed about having too many questions is: the ease with which people trim their question lists is correlated with their confidence in the project. In other words, hanging on to lots of questions is usually a sign of fear, uncertainty and doubt about a project. If you have fears about your project, don't puff up the project with too many questions. Improve your questions. How many questions people will answer meaningfully is a law of nature. As they say, you don't tug on Superman's cape, you don't spit into the wind, you don't pull the mask off the old Lone Ranger, and you don't mess around with cognitive budget. (Sort of.) (It's a song. Seventies. Swing set. Saturday mornings. After I swung up high enough, I could fling my shoes clear across the yard and onto the sidewalk.)

But duplicate questions are not always bad. Most of the time they result in wasted cognitive budget, irritated storytellers, and wasted sensemaking or analysis time. However, in two special cases they can be useful. The first case is where you know very little about your storytellers or expect excessive variety (say they are random people walking by a street corner). The second case is where you want to consider a particular topic intensely, say trust or creativity or decision making. In those cases, very similar questions can aid in mapping out an unexplored area of the conceptual landscape. For example, I've seen people ask both "How do you feel about this story?" and "How did this story turn out?" The answers are usually similar -- if I feel good the story turned out well -- but in the places where they don't match you can find useful pockets of unexpected insight. If it's important enough to spend some of your cognitive budget on it, spend away.

Happy question asking again! If anybody has comments or suggestions or questions or confusions or arguments about any of this book fodder (that's what it is, in case you didn't know; these are drafts for the second edition of Working with Stories) please drop me a line.

No comments: