“Hypotheses are what we lack the least.” - Henri Poincare
At Bridgeport Digital, we have a network of teams working together everyday, using their knowledge of Agile and Scrumwork to develop strategies that will help improve productivity and make our goals more attainable. It all begins with our method of hypothesis planning because without it, we would fail to bring our company exactly what it needs. Using the scientific method is a regular thing when you’re dealing with Agile and in this specific reading, we’ll talk about how we use it. Let’s start by defining ‘hypothesis planning.’
What is Hypothesis Planning?
Hypothesis planning is when we define the relationships between the variables we use to measure what can or what should be accomplished. Our strategy begins by identifying the goal (or outcome), the foundation (or assumption), the decisions (or hypotheses), and the outputs (or activities). We decide what our goal (or what output to anticipate) is based on the foundation (or assumption) we believe to be true, then we test our decision/hypothesis. We test out hypotheses because a hypothesis is:
“a proposition made as a basis for reasoning, without any assumption of its truth.”
Consequently, as we do planning, we base it on info that is current. Most of our decision making is based on assumptions. The assumptions we base on aren't always true. We need to identify the riskiest things we do when we plan and test those things out. As we do this, we go through a process called ‘risk reduction’ where we test the high risk items and we verify things that are true and not true. By doing this, we de-escalate the risk. Hypothesis planning should happen for customer validation while looking at productivity improvement; this is the engineering process we use at Bridgeport Digital.
Decision Latency And How It Relates to Hypotheses
We begin by identifying a goal we’d like to meet (a goal is a future state). Next, we identify our current state and the gap between our goal and our current state. We need to make decisions to move us from our current state to our future state. The decisions that we make will be based on assumptions. Assumptions are acted on as if they are true but we don’t genuinely know whether or not they are. Nonetheless, we have to make assumptions because often we don’t have all the information to feel (100%) confident in the decisions that we’re making.
However, if we wait until we have 100% of the information, we can’t make fast decisions. The best way to determine whether or not we were successful in our endeavors to reach the goal is through decision latency. Decision latency pertains to the amount of time it takes for a team to make a decision, in response to a change within the business. If we want to make decisions and be successful, we have to make decisions with less information than we may be comfortable with. Thus, I make assumptions because I don’t have all the information when we make decisions. Assumptions are the cornerstone to making decisions.
Now, decisions themselves have a life cycle. When I have zero to little information, we call this a ‘hypothesis.’ Then, when we start collecting information, we can form a ‘conclusion.’ Once the conclusion has been tested by our actions, we call that a decision. After the decision has been validated multiple times and we can use it to make predictions about the future, we call this a ‘theory.’ Once I have a theory that’s been tested multiple times, it becomes ‘knowledge’ used to make future decisions/hypotheses.
Formula for Executing Hypothesis Testing
For each part of the hypothesis, we can explore more. We ask what must be true to reach the goal (GIVEN), what outputs we need (IF), and the result we anticipate to obtain (by producing those outputs) (THEN). There is a relationship here:
GIVEN... we need _______________________.
IF... we want __________________________.
THEN... we will ________________________.
Ask, “What do I do to test it out?” Build the plan, execute it, and see if you're getting positive results. If you aren't, then summarize that it wasn't a good plan. The hypothesis was false and we should do something else.
Executing a Hypothesis Test Includes:
Building some of my plan, execute it, and measure if it moves me close to the goal.
Ask yourself:
What evidence do I and the people I work with need to see in order to validate the assumption?
Positive responses. People respond.
Other ways to ask:
What's a better indicator of whether or not the plan is actually helping achieve the results? How do we validate that the plan helped?
Experiencing Obstacles in Hypothesis Planning
You will encounter obstacles on occasion, in which case you report to your team and decide how to handle it accordingly. We refer to these as:
Impediments.
It can be difficult to break it down sometimes. This is critical thinking and this is what we try to teach other people how to do. People make mistakes because they aren't using critical thinking, our job is to help them learn how.
A hypothesis is a relationship between the assumptions, the actions, and the results we hope to achieve. If it's hard to prove the relationship you can choose a null hypothesis which is the opposite, to prove there is no relationship. But it's better to prove a relationship than to disprove one. If I try to prove my alternate or null hypothesis, I will achieve this result. Often when I'm trying to prove the alternate hypothesis and it proves to be true, I can leverage it for other operations in the future (I can reuse it). If you think your hypothesis is incorrect, you can choose the null hypothesis to save work.