Having an end goal in mind is sound advice, but charging forward without understanding what leads to what will waste time and resources.  Logic models are meant to outline assumptions, what resources are available, how those resources are being used, provide links from those actions to what is expected to result from them, and organize how those results fit inside end goal. Walking through a logic model prompts us to question exactly how each piece fits together. Logic models were originally designed with program evaluation in mind, but also serve as omni-tool in the design process as well.

What’s more, it forces us to answer what measures will actually be helpful in alerting us to what works, and what doesn’t. An example of a misaligned measure might be attendance in schools. If one knows the attendance numbers of a class, what does that actually tell us about the quality of the school? It could be that the school is extremely effective for those who are arrive, and that the attendence has more to do with the state of a community rather than interesting a class is. Knowing the number of students that arrived on a given day does not directly answer anything about the school or how well it is carrying out its activities. It provides context, but if my goal is for children to learn, then I should search for a more direct measure. Logic models can quickly give us this bird’s eye view of what our goals and intentions actually are, what to pay attention to, how to get there.

Before going further, the following shows a logic model of how just one aspect of a program was influencing its users. While the effort pays off in the end, working on models that attempt to showcase the logic of an entire program can quickly get out of hand.

 

Getting Started

To begin, my team created our columns: Resources available, Activities being done, Outputs, Short Term Outcomes, Intermediate Outcomes, and Long-Term Outcomes. Below our model, we included two additional sections: Assumptions, and Context. These are essential, and serve as a way to ground everyone in how a program is presenting itself, and what politics may be coming into play. Through brainstorm and research, organize your cards for each column. We don’t need to worry about organizing just yet. Sometimes working out just what the program activities or goals even are can take some time describe or agree upon.

It should be noted, that outputs are different than outcomes. Outputs are what we are measuring. An outcome is result of that action. For example, if my action is pouring coffee into a mug, the output would increased mug-captured coffee. The short-term outcome is a productive morning, which could have several other outputs contributing to it.

Putting it all Together

Now, it is time to begin making tentative pairings of what might be associated with what. This begins to inform what needs to be researched. Around this stage, there will be a few clear links that everyone can agree on, and some links that are less sturdy. Some in the group can be researching those clear links to ensure that there is solid literature to back up the logic, while others continue to refine placement.

I will not be obvious how some activities or outcomes eventually connect with each other. Sometimes working backwards from the final desired outcome can shift perspectives enough to keep creativity levels from falling, or getting stuck. This step can quickly become the most time consuming. Don’t be afraid to completely rearrange your cards and start over. That’s what they are for. Start over now, when the only thing to shuffle is cards and not people.

Adding Theory of Change

Once tentative research has been done, and everyone is feeling good about placement, we can begin connecting everything to each other. We used scraps of paper to showcase the theory of change along the program. This helps to uncover the driving force behind program outcomes, the path needed to get there, and what outputs can be measured to best understand the situation. By this time, our assumptions category may have grown. That’s fine. The logic model here represents the logic of how things change inside the box of those assumptions and context we listed earlier.

What to Measure?

These links are hypothesis that this card influences another card in a certain way. When laid out visually, we can see stress points, or areas where resource heavy activities. We can see if there is a general assumption that is present throughout the program journey, and we can see what aspects of the program are acting as interchanges where paths meet. The links between cards are great places to begin discussions on measurement. Having data on your program to back up that A does or does not leads to B could have a monumental impact on what services a program continues to fund, or brand itself.

Wrap-Up

1. Logic Models push us to be specific about are goals and the paths that lead to those goals.
2. Grab index cards and paper. Think carefully about what your assumptions and context are. Begin brain storming and sort cards into columns
   1. Resources available
   2. Activities being done
   3. Outputs
   4. Short Term Outcomes,
   5. Intermediate Outcomes
   6. Long-Term Outcomes
3. Divide and conquer. Have some teammates continue the discussion, while others fact check and research relationships of less complicated links. Do not be afraid to restart the entire model. It will happen.
4. Connect items together either working from resources to goals, or from goals back to resources. Use these links to see what is driving the program, and what links need to be true in order for the program to work. Evaluate the strength of those links.

Logic models can be used anywhere. Ours helped us role-play our stakeholders and select evaluative questions that would be valuable for them to know the answers to. In UX design, logic models can be used to see where assumptions are being made, what is leading to an expected reaction, and what user values may be.

Logic models can be a bit intense, and they take several versions before they are helpful. The reward is a streamlined understanding of what is going on, what the program is doing, and what it needs to do to be successful. Give it a shot! Experiment to see how creating a logic model can help you in your practice!

I love learning about games and the data that surrounds them. Often, large game platforms are surrounded by large numbers, and staggering comparisons (here’s one from Bungie regarding Halo 3).

This usually results in impressive stats in millions, billions, or trillions. Something I am curious about is how these numbers actually manifest in what the player thinks about a game. At some point, these data feel just a little hollow. Numbers can be insightful, but often leave out the voice and the relationship that we have with games. With all these numbers, how do players describe their experiences with these games?

I created 4 surveys for the purpose of exploring how players value their relationship with several games from Blizzard Entertainment: HearthStone, World of Warcraft, StarCraft, and Overwatch. I am primarily interested in understanding what, if any, commonalities exist between how players describe an “ideal” match or play session and what their favorite memory of that game is. What I hope to end with is a coding scheme that could be used to guide game and UX designers in crafting experiences that have the ingredients for being memorable.

The following quote, I think, is a perfect example of this, and why we need to understand how these games fit into people’s lives, and how games can leave us as better people for playing them. This is a response to the question “What is your favorite memory from playing World of Warcraft?”

“All the time I spent with the many friends I made there. They were some very formative years where I grew a lot as a person. For a long time the majority of my close friends were people I knew exclusively on World of Warcraft. I actually met my first boyfriend on there, who I dated for about half a year (which seemed like a long time as a 15-year-old). Every time I’ve taken breaks from WoW and come back, I always seem to find new people who I grow close to. I even met two of them in person. I can’t really choose just one memory, but truly those first two years with my first two groups of close friends are always a joy to remember.”

Starting with this qualitative approach allows responses to go above and beyond survey items which can indicate where future quantitative analysis should be focused. There is also a reason qualitative data is not often collected at the scale of millions, billions or trillions: it takes a long time to code. Over the next few weeks, I’ll be reviewing this data. Once done, I will update this post with the results.