The Grandfather Dialogues
How Do You Solve Problems, Grandfather?
by Larry Sawyer
IIA 1974

"A problem can't be truly solved without knowing the cause."
"Every problem?"
"Every problem."

How Do You Solve Problems, Grandfather? (Chapter Ten)

Grandfather, you once told me an internal auditor should think like a manager.

That's right, Randy.

Well, how does a manager think?

Do you mean a modern, professional manager?

Yes, of course.

He thinks with his head.

Stop joshing me, Grandfather, I'm asking a serious question.

I'm not joshing and I'm giving you a real serious answer.

You'd better spell it out for me.

Very well. Suppose your sister Karen asked you what you thought of her new dress - a really fine dress. What would you say?

I'd say it was awful.

Even though it was a very nice, well-made, well-fitting dress?

I'd say it was awful.

There you go. Here I tell you that by all standards - by anyone's standards - it is a superb dress, and all you can say is "it's awful." Were you really thinking with your head?

I guess not.

No, you were letting other things affect your judgment. You were not objective. You were not concerned with facts. You were shooting from the hip - not using your head.

But I'm just a teenager with normal opinions about his younger sister. Aren't managers grown-up thinking men?

Many believe they are. But they may not really think their problems through. They may make snap decisions, based on faulty experiences, guesses, intuition, business folklore, or what the last person said to them.

That's not so good. If an internal auditor thought like that kind of manager he'd be in real trouble.

He would, indeed.

How about a modern, scientific manager? How does he think and how does he solve problems?

We hope he'd solve his problems through a rational logical process.

What's that?

First, take a look at the manager of the future.

I don't see him. Where is he?

Okay, wise guy, just listen. The manager of the future - and many of the managers of today - will know less and less about the skills and the knowledge of the people they manage.


Because life is getting much too technical and complex to know or understand thoroughly everything that's going on.

What will they understand?

They will know how to manage the operating techniques of the people they manage - not what they do, but how they go about doing it. And most of all they have to be able to ask the right questions.

Is that important?

Decidedly. The thing which holds the modern manager above the crowd is his ability to ask the right questions. Particularly the questions that give him the answers he needs to help him plan and make correct decisions.


Because planning and decision-making are a manager's food and drink. That's what he lives on. That's what he gets paid for: to plan so that problems stay away from his door. But when problems arise, he must know how to solve them. And he must teach his people how to solve them. So, to solve problems he must be able to ask questions that search out the cause. A problem can't be truly solved without knowing the cause.

Every problem?

Every problem. Because any given problem has only one cause.


Yes sir. One event or combination of events that will always produce the problem - the unwanted effect.

Like in a laboratory?

Exactly. In the laboratory, every time you combine hydrogen and oxygen in the right amounts and under the right conditions, you produce water - the effect. Similarly, when you're faced with a problem you want to know what elements will invariably create the same effect.

Is that the way you solve problems?

Just about. Actually, there are four steps to problem solving.

What are they?

One - specify the problem. Two - analyze the problem to find possible causes. Three - test for the probable cause. Four - verify the most likely cause.

That sounds easy.

It isn't.

Why not?

Because each of those four steps has to be followed carefully and thoughtfully.

Say, that sounds just like a detective story.

You put your finger on it, Randy. It's like a whodunit. You have to search for clues by asking the right questions. You keep narrowing down the search until you pounce on the real culprit.

Can this be done scientifically, like in a laboratory?

I think so.


First you have to define your terms.

What does that mean?

You have to decide what the important words really mean.

Like what?

Okay. You tell me. What is a problem?

Well, it's a question you have to get an answer to.

That's much too general for our purposes.

Okay, smart Grandfather, you tell me.

Have respect, young feller. A problem - a business problem - can be looked at in this way: First, every problem is a turning away, a straying from something you'd expect. It is a deviation from some standard of expected performance. Second, every cause of a problem can be traced to some change.

Hey, Grandfather, you've got me dizzy.

Very well, let's break it down this way, in a situation that may actually have happened to you. You're riding your bicycle to school in the morning. It is running smoothly. Suddenly it starts riding roughly and you can feel the front wheel bumping on the pavement. What have you got?

I've got a flat tire.

You've got a problem.

You'd better believe it, Grandfather.

Pay attention, now. You've got a deviation from expected performance. The expected performance is a smooth ride. The deviation is a rough ride. Now what do you look for?

A nail in my tire.

No. You're jumping too fast. You're looking for the cause. And there may be many possible causes.

Like what?

Flat tire. Broken spokes. Out-of-round wheel. A suddenly bumpy road. Lots of possible causes.

I see. And now I must check for the probable cause, after I've looked at all the possible causes.

Great. Now what's your probable cause?

A nail in my tire.

You sound like a broken record. You're still jumping too fast.

I'm sorry. It's a flat tire.

Splendid. And now you're looking for the change that caused the flat tire.

Like what?

The change may have been the addition to your tire of a tack or a piece of glass or a crack in the tire's fabric.

Why do you make it so complicated, Grandfather? I got a flat because I ran over a nail.

I'm talking about principles that apply to any problem, and you keep talking about a snap conclusion that you made without even getting off your bicycle seat.

What's wrong with my conclusion?

The facts were that you felt a bumping of the front wheel. That could have been caused by anything. If you shoot from the hip instead of taking careful aim you may be way off. Let's get back to principles, even though the situation is a very simple one.

I'm sorry.

Forgiven. First, we had a deviation from standard performance. That could apply to a bicycle or to the stoppage of an aircraft assembly line. That is what is known as a problem - the deviation. By looking around and by asking questions we identify the problem.

Is that important?

Very. In the bicycle case it's simple. You take a look and you see you have a flat. In complex business situations, identifying the problem isn't that easy.

Okay. What's next?

Determine the change that caused the problem.

What kind of change?

A problem is a deviation from a standard. The change causes the deviation. Can you see what the change was from sitting on your bicycle seat?

Sure. My tire was round and now it's flat.

Wrong. The flat is part of the problem, not the cause. The change - the thing that would create the flat tire every time under similar circumstances - will be found only when you get down and examine the tire closely.

I just can't believe it. You take something simple and make something complicated out of it.

Okay, smart guy. Will you agree that to cure the problem you have to know the cause?

I agree.

Very well. Here's how the tire could have gotten flat.

  1. A tack or nail entered the tire and punctured it. A piece of broken glass gashed the tire.
  2. The valve stem came out.
  3. The valve stem broke or got jammed in an open position.
  4. Excessive air pressure and heat burst the tire.
  5. A twist in the tube pinched it.
  6. The smog ate it away.
  7. Your dog Popi may have been chewing on it.
I guess you're right, Grandfather.

Could you have identified each of these causes - changes - while sitting on the bicycle making snap decisions?


So you have to identify the particular change that caused the problem. Okay?


That's why I was trying to get you away from your particular flat tire so that I could show you the principles you use to solve just about any problem.

I see. It's just those words you were using that bothered me.

Sometimes it is better to think with a certain word so that you stay on track.

What do you mean?

If you ask yourself "What was the deviation?" you might get better answers than if you asked yourself "What's the problem?"

I'm beginning to get it.

And if you ask yourself "What was the change?" you might get better answers than if you asked yourself "What's the cause?"

It's like looking through a microscope instead of through a telescope, isn't it, Grandfather?

Well said, Randy. In both cases you're looking through lenses. But the microscope helps you focus on the precise detail.

Well, now that you show me how important principles are, can you give me some principles for analyzing problems?

I'll be glad to. And I want you to relate them to your flat tire.

I'll try.

Principle 1. To analyze a problem you need a standard - how things should be - to compare performance with.

On my bicycle, the should is a smooth ride.

Principle 2. A problem is a deviation - a departure - a straying away - from a standard of performance.

On my bicycle the smooth ride became bumpy.

Principle 3. A deviation must be precisely identified, located, and described. Look for the place, the time, and the extent. Where it happened, when it happened, and just how bad the situation is. Also, try to show what the problem did not include.

You'd better give me time to figure that one out. Let me see. As to where it happened, it was in the street on the way to school - not on a bumpy, junky lot, full of nails and broken glass.


As to when it happened, I was riding in the morning - not in the heat of the day when the tire pressure might expand and burst the tire.


As to how bad it was, the tire went flat, but the rest of the bike was okay.


As to what it didn't include, the rear tire was okay, but the front tire was not.

Splendid. Now, Principle 4. There is always something which stands out, which the cause has affected but which didn't affect anything else.

On my bike the tire was affected but the wheel and the spokes were not.

Principle 5. The cause is always a change. Something distinctive and different. Something that produces the new effect - the unwanted effect.

My tire was changed. Where it was nice and round it's flat as a pancake ... on the bottom of course.

Principle 6. Possible cause must be deduced - like a detective examining clues. One must find the change and then figure out all the possible things that could have caused it.

My tire could have gone flat because of any of the six things you talked about before. Like the nail or the glass or the valve stem or the heat or the twist in the tube.

Principle 7. The most likely cause of the deviation from the standard explains all the facts in the problem.

After I take a good look at the tire, I see a nail head sticking out of it and that explains it to me. And I don't stop looking when I find the nail head. I go over the whole tire.

Great. That's the line of reasoning that helps you find the cause.

That's an awful lot of work. All I had to do is get off my bike, spin the wheel, and look for the nail.

Sure. In a simple problem like that it's quite easy. But when you have a difficult business problem it isn't that easy. If you don't define the problem exactly, if you don't hunt for the changes methodically, you might jump to erroneous conclusions and take expensive corrective action that never corrects the real problem. You'd be shooting from the hip, not using your head.

Okay. But even in business problems, do you have to go through the whole routine?

Not always. Simple problems can be attacked quickly and the deductive steps are taken in a flash. But the smart manager follows an orderly process - even if it's a quick one - on most of his problems.

Do you mean that all smart managers use this system?

Oh, no. There are many different systems. I'm just partial to this one. But every good system involves an orderly process of defining problems and looking for causes.

What do you do after you've found the cause?

Then you make a decision.

I know that you've got all kinds of meanings for simple words. So you'd better tell me what you mean by decision.

It's a choice among various ways of getting something done.

Do you need principles for those too?

You bet. Shall we go over them?

Okay. And I'll try to tie them to my bicycle problem.

Principle 1. First establish the objectives of the decision. Just what do you want to accomplish?

That's easy. Make my tire round again.

That proves you haven't really defined your problem. That shows how failing to define your problem leads you to faulty actions.

There you go making things difficult again.

Now just hold on. Remember that I told you that you were between home and school on a bicycle with a flat tire. Now what are your objectives?

I read you loud and clear, Grandfather. Here are my objectives: Have my tire fixed, keep my bike from getting stolen, of course, and arrive at school on time.

That's better. Principle 2. The objectives are classified as to importance.

Let me think. My bike is worth $60. So keeping it from being stolen is my first objective. Then, I'd better get to school on time, because the tire can really wait. Down at the bottom, I guess, is having the tire fixed.

Absolutely splendid. Now do you see how the organized approach to a problem gets you on the right track?

You bet. I'll have to teach this to my dog Popi. He makes problems all over our living room rug.

Very funny. Principle 3. Develop alternative plans of action. Set up different choices from which to choose the best one.

I can leave the bike where it got the flat and go on to school. I could ask for a ride from a passing car and have him take me and my bike. I could push the bike to the garage, about a half a mile away, to have it fixed. I could push the bike back home and run to school or have my Mom drive me.

That's quite sufficient for our purposes. Principle 4. The alternatives are compared with the objectives.

Okay. The first one is dumb. If my most important objective is to save my bike I shouldn't leave it in the street. Scratch that one. As to hitching a ride, my Dad said that I should never take rides from strangers. They might run off with me and my bike. Scratch that one too. I could push the bike to the garage, a half mile away. That would keep the bike safe and get the tire fixed, but I wouldn't get to school on time. I could push the bike home, so that it would be safe. I could then run to school and maybe not be late at all or be just a little late. Then after school Dad would help me fix it.

So far so good. Principle 5. The choice of the alternative best able to achieve the objectives is the tentative decision - the one that seems to be the best decision.

Okay. My last choice - to roll the bike home - seems to be the best one.

Principle 6. The tentative decision is looked at very hard for bad side effects - future possible adverse consequences.

Well, I might be quite late for school. I might get hit by a car while I'm pushing the bike. I might get too tired to run to school.

Well done. Now the last one. Principle 7. Be fully aware of adverse consequences and keep them from becoming problems of their own. Then make sure the actions decided on are carried out.

Okay. I don't think those adverse whatevers are so bad. I think I can make school on time, because I'm usually very early. I'll push my bike on the sidewalk so a car won't hit me. The bike isn't heavy, so I won't get too tired.

Very good, so far. Now how about making sure your actions get carried out?

The only thing I have to worry about is getting the tire fixed. When I get home I'll ask my Mom to buy a tire patch kit. Then after school I'll run home and clean up Dad's work bench to get him in a good mood. Then I'll ask him to help me fix my tire.

Wonderful. Now you've seen how this system works on a simple problem. It will work on difficult problems also.

Do these rules apply to internal auditors too Grandfather?

Absolutely. Whenever an auditor finds something wrong and makes a recommendation, he too should use his head and not shoot from the hip.

What do you mean?

Well, when he finds something wrong, he should be certain he has really and truly identified the problem. He should dig deep until he knows what the causes are. He should make recommendations that are reasonable, practical, and economical.

Could you set out some rules that the internal auditor should follow when he comes up with a deficiency finding and makes a recommendation?

Certainly. Here they are:
  1. Recognize the problem. Look for the deviation between what should be and what is.
  2. Specify the problem - the deviation - the deficiency - as precisely as possible:
    a. What is the deviation?
    b. Where is the deviation?
    c. When did the deviation occur?
    d. How big is the deviation?
  3. Distinguish between what the deviation is and what it is not. (It was the tire. It was not the rim.)
  4. Test for causes. Look for causes that completely explain the deviation, that would make the deviation happen every time, that answer all the parts of the deviation.
  5. Set out your objectives - exactly what you're trying to achieve with your recommended corrective action.
  6. Set out alternative actions. List them in the order of importance. Some may be musts. Some may only be wants.
  7. Compare alternative actions with objectives. Tentatively select the best.
  8. Think out the possible adverse circumstances - the bad side effects - the potential problems that might affect the tentatively suggested action.
  9. Think of contingency actions. Think of the "what ifs." Keep in mind another action if the first one won't work.
  10. Ask management to set up controls which will see that the action is really carried out. They can't just issue instructions to people and forget it. They must have some kind of follow-up mechanism to be positive that the corrective action will be completed.
So anybody - managers, internal auditors, and kids in school - can use this system to solve problems, right, Grandfather?

I think so.

Can you use it on a real internal audit problem?

Of course.

Is it like a detective story?

You bet.

Tell me about one, Grandfather.

Okay. An auditor told me about it. But it was an interesting case. He was making an audit of control over standard tools.

What are they?

Tools you can use on anything: Wrenches, saws, electric hand drills, and the like. In this case it was mainly electric hand drills, like your Dad uses in his work shop.

His cost $39.95.

Well, these were industrial drills and cost about $110 each. This auditor's company uses a great many of them; so they have a big investment in hand drills.

Who uses them?

The workers. The hourly workers. They're called that because they get paid by the hour. That distinguishes them from the salaried workers who get paid by the week.

How do they get the tools?

Each group of workers is supervised by a salaried supervisor. He goes to a tool crib and signs out for a whole bunch of the hand drills. Then he keeps them in a cabinet near where the workers are doing their jobs.

What does the cabinet look like?

Something about the height, width, and depth of your Mom's laundry closet. But instead of flat shelves, it has a lot of cubby holes. There is a drill in each cubby hole. When an hourly worker needs one, he goes to the supervisor and gives him a tool check for the drill.

What's a tool check?

It's a round piece of plastic with a hole in it and with the worker's clock number or name stenciled on it. The worker gets about ten of them when he gets hired. When he leaves the company he must turn in all ten or pay for each one he can't turn in.

So he is careful about them, isn't he?

He sure is. When he surrenders a tool check for a tool, the tool check is hung on a hook in the little cubby hole in the cabinet to show that it was given in place of the tool.

So what did the auditor find wrong?

He made an audit of the tool cabinets. There were about 250 of them. Each had about 25 cubby holes. Each cubby hole has a tool costing in the neighborhood of $100. So we're talking about $625,000.


Wow, indeed.

So what was wrong.

We found that almost none of the cabinets were kept locked. We checked a sample of 50 of them and found that almost all of them had one or more tools missing ... tools that couldn't be accounted for by a tool check. The situation was pretty bad.

How long had it been going on?

You're beginning to ask good questions, Randy. It had started since our last audit, two years ago. At that time we found that the cabinets were in good shape and were well controlled.

So there must have been a change since then.

A perceptive comment, my boy. Yes there was. But let me tell you how we worked it out scientifically. We wanted answers to four questions. Do you remember what they were?

I'll never forget: What, where, when, and how much.

Good. And how do we break each of the questions down?

Between what the problem - excuse me - the deviation was and was not.

Brilliant. Now look at this matrix -

This what?

Matrix. It's a layout of statements, words, numbers, or whatever, so that it is easier to see them all at once.

Okay, where's the matrix?

Here it is:

Specification of the Problem
Question Is Is Not
What Standard tools are missing Nonstandard tools
Where From tool cabinets From cribs
When Since the last audit. Going on for about two years. Before the last audit. Before two years ago.
Extent Most cabinets Cabinets under control of especially competent supervisors

By asking questions and doing some routine checking, the auditor was able to get the answers to those questions.

What did he do then?

You tell me. What do we look for now?

Beats me.

If you remember, after we've specified the problem, we then look for -

Oh, yes. I remember. You look for the cause.

How do you look for that?

You look for, er, er, extinctions.


Right. For what was different. And then you look for the change - something that explains all the facts in the problem.


What was the change, Grandfather?

It was really quite simple after he laid out his matrix and started asking questions from people in the factory and in Personnel.

What's Personnel got to do with it?

They take care of the paper work when people join the company, leave the company, or move from one department to another.

Why would an auditor even think of Personnel when he's worrying about tools?

It's an obvious connection, when the question of charging employees for anything because any decision to hold employees accountable is a Personnel responsibility. And two years ago, they changed the forms they used when people leave the company or move from one department to another.

What happened?

The old form had a place where both hourly and salaried people got clearance for all tools charged to them. The new form required clearance only for the hourly people.


Somebody goofed.

What happened?

Hourly people still had to get clearance. So there was no problem with them. But whenever salaried supervisors were responsible for tools, they knew that they wouldn't be held accountable. They wouldn't have to get clearances. So they left the cabinets unlocked, never looked for missing tools, and just plain neglected things. Some of the real good supervisors, who did everything else well, kept their cabinets under tight control. But the others did not.

But what about the hourly workers' tool checks that they gave to the supervisors for the drills and which got hung on the hooks in the cabinets?

Since the cabinets were left unlocked, it wouldn't be hard for a worker to take his tool check back any time he wanted to, if he lost or mislaid a tool for which he was responsible. Then when he left the company or moved on to another department, he'd be able to turn in all his tool checks.

What was the corrective action, Grandfather?

Very well, young man, since you now know all about decision making, you tell me. I'll go through the principles and you give me the answers. First, what are the objectives?

To keep tools from getting lost. I can't think of any others.

The one you have stated is right, of course. But we'd also be concerned with having the tools readily available for the workers. Then again, the tools should be on hand to keep them in good repair. If they can't be found they can't be repaired - and poor tools do poor work. Now, Randy, let's classify them as to importance.

Okay, Grandfather. First keep the tools safe. There's a lot of money tied up there. Then have them available. Then keep them in repair.

Some people might argue with you. But for our purposes let's leave it that way. Now for Principle 3. Develop alternative plans of action.

Let's see. We should change those personnel forms. Also, maybe we could send all the tools back to the cribs to keep them safe. And maybe the supervisors need to be told how to take care of cabinets.

Sheer genius, Randy. Now let's compare alternatives with objectives.

Changing the form, to make supervisors responsible again, seems like a good idea. It ties in with all the objectives.


We could send the tools back to the cribs and close up the cabinets altogether.

Hold on, Randy. That runs counter to the idea of availability of tools. If you didn't have cabinets people would have to wait in lines at the cribs. That would hurt production schedules and production costs.

I see. I also see that comparing choices with objectives keeps you on track.

Quite right. That's principle 4. But the supervisors should be told how to take care of cabinets.

Right. Now to principle 5. The best alternatives should be chosen.

I think we've already done that.

I agree. Now for principle 6. Adverse consequences.

We would have had some if we sent the tools back to the cribs. So we avoided that one. But here's another. If the supervisors are charged for tools lost, that would cost them a lot of money. Do you really mean to do that?

Perhaps not. That's a bad consequence. Maybe they won't get charged. But they should be reprimanded. And that would keep them on the ball. Now,the final principle. How do you make sure that the supervisors really do their job?

Maybe the managers should watch them more carefully.

Quite right, Randy. And maybe the supervisors should make periodic reports to the managers on lost tools and on the condition of the cabinets. That would make sure that corrective action continued to be in effect.

Grandfather, was that a real internal audit problem?

Indeed it was.

It sure was a lot of fun to work out the problems and the causes and the solutions with those principles of yours. And I know something else. If I have a flat tire I'll know exactly what to do.

Not necessarily.

Why not?

Circumstances may be different. Each deviation must be considered on its merits in terms of what, where, when, and how much. If you rely too much on precedent - on what happened before - you may be shooting from the hip when you should be using your head.

Okay, I'm going to start thinking about my main problem.

What's that?


You leave that lovely girl alone. She's my grandchild too, you know. So you'd better be nice to her.

Okay, Grandfather. Just for you I'll go home and tell her she's wearing a nice dress.

That'll be the day.

Audit Wisdom