Note: This article is reprinted with permission
from Jossey Bass/Pfeiffer from the forthcoming book, The Trainer's Portable Mentor of Training & Development.
Asking the right questions is much more important than having the
How can you ask questions that help you to assess training needs
when you don't know what you don't know; when you can not be sure
what the problem is? The answer is to use diagnostic questions.
These questions are different from interrogation or
cross-examination questions, which are designed to judge or assign
blame. Diagnostic questions do not assume the nature of the problem
is known, and they allow the people answering the questions to
guide the inquiry in the early stages.
This sounds simple, and for some it is. For most of us though, it
is difficult to ask questions that are not tainted with our own
preconceptions about what the problem is and, also, what sort of
solutions would be appropriate. Further, our perceptions of the
problem are likely to be inaccurate due to the biases that affect
how we attribute a cause of success or failure. Such "attributional
biases" affect all of us, as explained below. In sum, the
combination of jumping to a too-hasty diagnosis of the problem,
along with the attributional biases always at work when we perceive
causes of failure, can make performing an accurate needs
assessments a tricky business.
The difficulty of composing diagnostic questions is demonstrated
with the following brief case example. Students engaged in
role-playing this case example were instructed to generate
diagnostic questions. Below you will see some of their results.
Case example, Part A
"The Forklifts" case example is based on a true incident: A company
was experiencing problems because forklifts were breaking down much
more often than expected. Executives had evidence that the forklift
operators were not performing routine maintenance on the machines,
maintenance that would have prevented many breakdowns. It was
assumed the operators needed further training in maintenance
procedures. The company called in some consultants.
The forklift case example was given to a group of junior and senior
undergraduate students in my management development and training
class. In teams, the students prepared questions to ask of other
students who took the role of forklift operators. Those acting as
operators were told the "whole" story (you will see this in Part B
below), advised that they did not want to appear to be at fault in
this, and told to otherwise play the role as they chose.
Following are some of the questions students asked in performing
their diagnosis. After each question I have noted the unwarranted
assumption embedded in the question. The questions are arranged in
two groups. The first five questions are judging/blaming questions,
ordered worst to least bad. Following those are six examples of
better diagnostic questions, relatively free of judging and
Please note that during the in-class exercise, students were only
able to come up with one exploratory diagnostic question (the last
one listed, regarding improving working conditions). Fun exercises
to improve ability to ask assumption-free questions are the old
puzzlers like this one: "Dick and Jane lay dead on the floor. The
curtain was flapping in the breeze. The door to the room was
locked. How did Dick and Jane die?" Then allow questions to which
the only answers can be either yes or no. (The answer to the riddle
is at the end of this article.)
Judging and blaming questions
- Why do you think proper maintenance is not being done on
forklifts? (Assumes improper maintenance is the problem)
- What can you tell me about the forklifts breaking down?
(Assumes breakdowns are the problem)
- Why do you think the forklifts are breaking down? (Assumes
these are breakdowns [not sabotage] AND that this person knows
something about the breakdowns and sees the breakdowns as a
- Do you think we have a problem with the forklifts? (Assumes the
forklifts are the problem)
- Tell me about the problems you have at work. (Assumes there is
awareness of a problem)
Now let's look at some questions that will allow the necessary
information to emerge.
Exploratory judgment-free and blame-free questions
- Tell me about your work.
- Tell me a story about your best day at work.
- Tell me about a worst day at work.
- What questions should I be asking about your work?
- What should I understand about your work?
- What are some suggestions you can make that might improve your
Note: Since exploratory questions are so very open, they usually
should be followed by questions like, Could you tell me more about
that? Could you give me an example?
Case example, Part B
The consultants called in did not assume that the nature of the
problem was known. Instead, they conducted a full-scale diagnosis
of the problem. They engaged in face-to-face personal interviews
with a horizontal as well as vertical cross-section of employees.
They fed back their data in a "mirroring" phase. For example, if
researchers found that one of the themes to emerge from the
interviews was the lack of communication and participation in
decision making, they would identify this theme and then cite
several verbatim statements made by (unidentified, anonymous)
employees. The "mirroring" session would provide these detailed,
verbatim statements, such as the following: "I don't know why the
bosses want us to unload those Wednesday trucks way back in the
worst area of the parking lot. I think there are lots better places
and times that would make my unloading job easier. But no one asked
me about it, so I just do what I am told."
The consultants discovered that the combination of driving the
forklifts on uneven ground outside the warehouse and uncomfortable
seats without springs was causing physical distress to the
operators. The operators began to hate the machines, and due to
that hatred, they chose not to perform the prescribed maintenance.
They knew how, they knew why, but they would not do it. This is a
case in which a cursory diagnosis might have led to futile
training. Instead, the consultants uncovered the true problem,
which did not require a training intervention.
Five principles for conducting training needs
Principle 1: Do not assume the given or apparent problem is
Move around the problem to ask questions from those above, below,
in front, behind, and in the middle of the problem. Avoid questions
that are problem-focused or solution-focused, because this often
limits your inquiry into the nature of the problem.
Principle 2: Ask exploratory questions.
Do ask people to describe, to explain, to tell you about their
work. Do use critical-incident-type questions through which they
are asked to describe an especially good time at work, or an
especially bad situation. Feed back your data to the people you
interviewed to jointly interpret the findings.
Principle 3: Examine the reward systems surrounding the
Consider both the formal and informal reward systems in the
organization. Here, the classic article "On the Folly of Rewarding
A While Hoping for B" is perennially relevant. Are there any
rewards for desired performance? Or are people either not rewarded,
or perhaps even penalized, for performing as desired?
A former student told me the story of the military in his country
being concerned about the high cost of weapon repairs. The country
was not at war at the time, so they decided to penalize officers
whose troops had high repair costs. The officers quickly discovered
that the more the weapons were handled and used by the troops, the
higher the repair costs. Their solution? Lock away all weapons, to
be taken out only when ordered. The troops rarely handled their
weapons, and repair expenses dropped, but how well-prepared were
these soldiers? It was a classic case of rewarding efficiency at
the expense of effectiveness.
Bringing this back to our forklift case example, what happens when
a forklift breaks down? Does this mean the operator has a break
from work while a repair or replacement is completed? What behavior
Principle 4: Be aware of attributional biases.
Attribution theory tells us that we all have a tendency to protect
our own egos when we fail at something, by attributing our failures
to factors external to ourselves. Thus we fail to achieve our sales
quota due to a slump in the economy, problems with the product, bad
luck, etc. However, when we observe another person's failure, we
tend to attribute it to factors internal to that person. So when
our coworker fails to achieve his or her sales quota, we tend to
believe it is due to lack of ability or effort. This is why the
workplace is full of so many stupid and lazy people. Of course, we
ourselves are not like "them."
When I am observing others, I will tend to attribute failures to
the lack of motivation or skill level of the people I am
interviewing about the problem. Conversely, the people experiencing
the problem will be affected by the self-serving bias that leads
them to blame external factors for their problems, not seeing their
own deficiencies in ability or effort. How can we avoid or correct
for these attributional biases?
First, avoid the appearance of blaming. Often there is a tendency
to point fingers and to focus on identifying the person or persons
"causing" the problem. We assume that, if there is a problem, there
must be someone causing it.
One helpful strategy to circumvent this cycle of blaming and
denials is to conceptualize the problem as distinct from any
person. Make this problem an actual character in your story. In the
forklift case, we might have "The Breakdown Demon."
Principle 5: Allow for the unexpected.
Do ask people to suggest to you what else you should be asking
about. Let the interviewees guide your inquiry in the early stages.
Be open to people mentioning something totally unanticipated, and
following up on it. Pay attention to non-verbals and supposedly
irrelevant side comments, and use open follow-up questions to
explore these leads.
The first step in training needs assessment is awareness of
attributional biases that lead us to blame another's lack of
ability or effort for failures. Armed with such awareness, we can
better design relatively assumption-free questions. Asking true
diagnostic questions that are exploratory rather than judgmental or
blaming creates an adventure of discovery, and is the basis for an
effective needs assessment.
So how did Dick and Jane die? The wind killed them. They were
smothered to death in the air when the breeze blew their bowl off
the windowsill. Dick and Jane were goldfish.