Agile Project Management Viewed from Behavioral Science

Background
Earlier this year I attended an interesting talk by Tony Parrottino on Applied Behavior Analysis Science and posted a short write up. Recently I have had a couple of follow-on meetings with Tony and have become fascinated by the science of behaviour. I think it is a powerful, but under utilized resource for project managers, yet I am still trying to reconcile it with some other theories. Fortunately, Tony agreed to help outline some of the key concepts and clear up some issues.

1) Mike: Q: As a project manager I spend lots of time managing budgets and schedules, but in you say that in reality we can only hope to manage behaviour. Can you explain this?

1) Tony: A: Sure. Simply put, the “means” to all business results is “behavior” or what people do. That’s why we hire them – to produce these results. Money (budgets) for example is usually a “measure” of a result (selling value for example) or a behavior (labor cost for example). It’s not a result – we don’t actually make it, unless you’re a Counterfeiter? So if you think about it more precisely you really are “managing behavior” and the budget is really one “measure” of how well you are doing that. The same can be said about “schedules” or time in terms of managing a person’s behavior. If you’re doing that well the measure of “time” or what we call one dimension of productivity in business should improve. This is not a question of semantics either. This is a question about scientific precision and understanding what variables you have control over as a Project Manager – like behavior. Behavior Analysis Science helps managers understand how to “control” performer’s behaviors to positively enhance things like productivity and budgets. 

2) Q: Following an agile methodology approach, our team meets daily to briefly review work and issues. Everyone answers three questions: 1) What have you done since yesterday? What do you plan on working on today? What blockers or impediments do you have to progress? I recall you thought this was a sub-optimal set of questions; can you explain why, and suggest a better set?

2) A: Well Mike I apologize if I said these questions were “sub-optimal” – I didn’t mean to be so critical. This is another great question but requires an extremely lengthily answer to be complete, as it has several elements related to performance. So I will be general and brief in my answer and hope to make just a few points here. My first comment may sound strange but here it goes: “managers” tend to “over-emphasize” “asking questions of their performers” and also focusing on “removing” impediments. At the heart of performance management is the focus on not what people typically have to say, but what we want people to accomplish precisely and what we want people to do precisely. When we start there the data provided gives us objective information about how well any individual or team is performing or progressing. Also, when we focus on what we want and reinforce it we simply get it. If managers are finding they are spending a lot of time managing what they don’t want (like impediments, blockers, etc.) this is a strong sign of managing what we call “a deviation of performance”. This may not be obvious to most but I’m sure that every project manager has experienced a lot of spent time trying to manage what he doesn’t want? Just remember trying to remove what you don’t want will not ensure you will get what you do want! I would suggest if you feel there is value in asking daily questions to start with what precisely you want from your performers (results data preferably) and have clear objective measurable data on that performance. Then, here is one of my favorite questions I like to ask, after reviewing the data for progress; “How did you do that?” In this way you can reinforce the behaviors to accelerate the performance further. This may seem somewhat strange to most managers as most are constantly in problem solving mode and want to “remove” obstacles. If you find yourself doing this often I would recommend a review of your pinpoints and make sure they focus on what you want.

3) Q: So, what is Pinpointing and why is it important?

3) A:  Pinpointing is a technique that simply means “being precise”. When we want to manage anybody’s performance we need to be precise about two very important variables – results and behaviors, and in that order. These variables define what we want from our performers. Even when measuring these variables we seek the technique of Pinpointing to be precise about the measures as well. The importance of pinpointing should be relatively obvious. Take Project Management for example. If you’ve every asked someone or even yourself “I don’t think that’s what I wanted, or asked for?” I’m quite sure you haven’t pinpointed. I’m always amazed at the lack of “performance precision” that occurs in most organizations. Although pinpointing defines performance precisely, don’t let that fool you – it’s only your first step in successfully managing your team’s performance. Just because you’re precise about what you want doesn’t mean your going to get it now does it? 

4) Q: OK so how would you go about pinpointing for the common roles on software development projects, namely Project Managers, Software developers, Business Analysts and Quality Assurance roles?

4) A: Well the answer to that question might be too lengthily for the space we have here, as the analysis and development process of Pinpointing is not an easy technique. However, I will give you some insights as to where I would start. Firstly, the place I would start is not with any of the job roles at all. I would start with the Project itself. And I would ask the most fundamental question of “what is the one mission critical result” of the Project? Now before you start thinking too hard to try and postulate an answer, keep in mind that most people don’t know the difference between a result and a behavior (activities in this case as a Project doesn’t have arms and legs so it can’t behave). Let me make a few points here. Firstly, when we pinpoint we should always start at the highest level relative to where we stand “organizationally” before we get down to the job roles. Then, when we have that figured out if the next level is job roles once again we should start and only focus on the results for those job roles. This process or technique Tom Gilbert, a performance expert, called “mission pinpointing”. And the critical point to be made here is to never focus on behaviors until you have the “right” result, and even then don’t worry about behaviors until you have the “right” measures of those mission critical results. There are several tests and methods to not only determine pinpoints but also to know if you have a pinpoint. In short, I would recommend starting by taking a training course on “Pinpointing”.

5) Q: I recall you pulling me up on a comment that I made about the IT industry. I thought it overly biased towards left-brain analytical type people who might be less likely to be looking for people-based solutions to project challenges. We discussed the whole validity of the idea of personality types, which you discredited. Can you explain why you believe Myers-Briggs, DISC, Belbin, etc tests are flawed and what you would suggest we concentrate on instead?

5) A: The tests you mention are fundamentally NOT rooted in any science. The subject matter of “behavior” which is of critical importance to any business manager, as it is the means to all business results is fundamentally what we have been discussing. The question to be asked on these “tests” is “what use does the information provided by them present to a manager”? Generalizing, grouping, organizing, and cataloging “behaviors” may have a place in informal coffee discussions but really provides NO value in “explaining” the precise cause-effect relationships of the important variables related to “behavior”. You can’t honestly believe that by simply answering a set of test questions you could explain all the detailed complex and functional processes of “behavior” and furthermore interpret it, predict it, and control it – the use to a manager? Behavior Analysis Science has almost one hundred years of refined scientific and experimental study. Science is the closest thing we have to the truth on any subject matter. Knowing that there is a science to “behavior” is your first clue on what to concentrate on instead of explanatory fictional tests.    

6) Q: Is Applied Behavior Analysis science or psychology?

6) A: Science. Ironically, it got its start in psychology. When Pavlov discovered the process of “reflex conditioning” the “psychology space” started to “extend” the functional relationships of stimulus-response to try and explain all “behavior”. This proved to be futile. But let’s back up a second. Pavlov himself was a Physiologist and was studying the scientific process of the digestive system. Serendipitously, and very scientifically he measured, observed, and discovered the causal variables related to reflexes and “reflex conditioning”. It wasn’t until Thorndike, a psychologist, started to study, measure, observe and document the effect of “consequences” on learning in animals that non-reflex behavioral relationships “started” to take a somewhat “scientific” understanding. Finally, the person that built on and refined all of these scientific studies of behavior was B.F. Skinner. That’s when fundamentally the studies of “behavior” diverged into Psychologies and Behavior Analysis Science. Behavior Analysis Science discards “inner” variables as causes of behavior. For example, how would you measure, observe and explain the functional relationship to behavior of an “ego”? Other sciences like Physics, Biology, and Chemistry rely on physical, measurable, and observable variables to explain how things work. And like these sciences there is only one science of behavior, Behavior Analysis Science. I’ve quite frankly lost count as to how many Psychologies there are?    

7) Q: You emphasize the importance of effective human performance measurement. What should we be looking for? What are the characteristics of a good performance measurement metric?

7) A: This is a great question because measuring human performance has been thought of as a “mystery” to most managers. That’s because “we” think that human performance is some “internal” process that can’t be measured. This is fundamentally false. Human performance is an “outside job”. Its real measures are in what we accomplish (results) and what we do (behaviors). It’s that pragmatic. When we start to look at the real world variables of results and behaviors we can now not only measure human performance we can observe it precisely as well. Keep in mind that the real business “value” is in the result. The behaviors, even though they are the means to the results, quite frankly, “cost” the business – time, effort, money. Often, we spend a lot of time measuring and observing what people do (behaviors) and believe we can measure performance. This is very misleading, as the real value of performance is in the results. In other words “don’t confuse the plough with the crop”. A characteristic of a good performance measurement metric lies in the heart of “pinpointing” again. The more you are precise about what you want to measure the easier it is to measure. If you’re precise about the result you want for example then measuring it is a simple exercise in counting. For example if you know you want precisely 2”X4”X6” pieces of wood (quality measure) and you know you want 50 pieces a day (productivity measure), your ready to measure the precise specifications of the result (wood) – you merely count it. This external focus and pinpointing approach also works for “knowledge” workers. We don’t hire people for their knowledge; we hire people for their results from their behaviors which demonstrate their knowledge.

 

Finally, where can people find out more information on Applied Behavior Analysis Science?

If you’re interested in applications to business performance I highly recommend two books:

1. Performance Management - Aubrey Daniels and James Daniels
2. Human Competence – Tom Gilbert (1976 publication)

If you’re interested in the actual science of behavior I highly recommend this free pdf book: http://www.bfskinner.org/BFSkinner/PDFBooks.html 

And of course anyone who reads this article can feel free to call me directly for more information at 403 461 9100 in Calgary, Alberta, Canada or visit my company’s website (AnMar Management Inc.)

Thank you Tony for sharing your insights into behaviours on teams.

Comments

[billy - los angeles project manager]

I was continually frustrated with managing Graphic designers and Web Programmers - who had no idea of time management and budgets. Agile project management is something very interesting - will explore further - thanks!

[Dmitry Zdanovich]

IMO, it's too categorical opinion.
How to measure developers' performance? Lines of codes? Come on.
How to measure performance of individuals on the team? Their impact on the team's performance?

And if you say to your subordinates that you want something be done, it doesn't mean that they do that really good. They can even did it formally according to your precise measures, but at the end of the day it will hurt overall result.

Sometimes result is too multi-dimensional thing to be measured simply. Although there are some areas where similar approaches can be used.

[FRANK SETO]

The connection between Social Sciences like Psychology and computer design is blooming.
Check out the University of Ohio for more details, eh?
Frank Seto in Calgary
http://ifranks.blogspot.com

[Mike]

Interesting stuff (despite the need for a proof-reader) and I like the simplicity of the focus on results and behaviour. I agree that those personality type tests are flawed. I think they're no more scientific than the quizzes in teen magazines. Also it's very true that people don't understand the difference between a result (a noun) and an activity (a verb). It would be helpful if MS Project would explicitly differentiate the two.

[Term papers]

Thanks for sharing this information and hope to read more from you.