Welcome to Wednesday night! Tonight we have a few ideas to discuss, and then I’ve planned for us to use the second part of tonight to get feedback from our colleagues, from other teams.

Let’s continue to integrate all the ideas we’re bringing together in this class.

And here’s how we’ve been practicing the ideas we’ve discussed so far.

Up next, you’ll be presenting to the CEO of your organization, you’re presenting use your project as a way to demonstrate value and convince the CEO to invest further in analytics.

I’ve also given you an extension of your interactive communication, which is now due at the same time as your presentation. The idea is to use that additional time for feedback and iteratively improve.

Ok, during the semester, as we’ve thought about new ideas and examples of those ideas, we’ve tried to generalize from examples, and to think of examples of generalizations. Next, I’ve like to systematically categorize the kinds of things we want to look for, question, and answer as we review data-driven communications, as well as have a process of working with others.

We want to consider four things when asking help from others.

First, the person asking should establish the purpose of the critique. Be specific.

And the person reviewing should, second, be objective and think through your reasoning, third, offer alternative solutions to issues you identify, and fourth, structure the review. By structure, I mean start with a global assessment. Point out issues with alternative solutions, and also note what works well in the communication.

Now, we’ve discussed many ideas in terms of what to look for. Let’s review some of these now.

First, and foremost, we must. We must identify the audience and purpose of the communication. Don’t bother with anything else until these are specifically identified.

Next, review the messages, the narrative. How do the ideas link together?

Does the narrative use comparisons and show change?

For the data visuals, how are they encoded. Do they apply each of the principles we’ve discussed?

What about hue, saturation, and luminance? How are each of those channels of color used. If at all? Is there an intentional purpose for each use?

Have optimal annotations be layered into those encodings?

All of these ideas we’ve discussed throughout the semester, right?

Systematically review the commmunication against these ideas? Do they follow the advice? If so, how? If not, would it be more effective if they did?

What we’re working on in here is building awareness to help you learn to critically see all aspects that affect your communications with data.

That remains true for presentations as communications, too.

And you’ll be presenting next week. So let’s discuss presentations specifically. Let’s consider a few ideas that may help you plan for them.

And I’d like to start with a critique of them. Edward Tufte says, and I’ll read this with you:

PowerPoint, compared to other common presentation tools, reduces the analytical quality of serious presentations of evidence.

This is especially the case for the PowerPoint ready-made templates, which corrupt statistical reasoning, and often weaken verbal and spatial thinking.

What does this mean to you? Should we not present using powerpoint and similar tools for presentations?

Tell you what, let’s see whether Tufte does.

I’ve pulled this youTube video of a Keynote Presentation by Tufte more recently, in 2016, at Microsoft’s machine learning conference. It’s an interesting presentation and you should watch it.

But for tonight, I just want to see whether Tufte has presented without using powerpoint like visuals. Starting play will jump to the middle of the presentation, let’s take a quick look.

You can see an audience in the foreground, right? And Tufte in the background to the right. And left of him, or rather, to his right, we see he’s using presentation sliddes.

So he must not mean never to use them, right?

But let’s keep his critique in mind.

Now, I’ve asked you to review presentation material from several sources. Compare those you’ve seen in this course,

with the way Jonathan Schwabish recommends creating visuals, and the way he’s presented.

Then, Look for similarities and differences between those and the presentations by a couple of data visualization influencers. Here’s a screenshot of several presentations by Lisa Charlotte Muth.

And here’s a few screenshots from another influential data visualization expert, Nadi Bremer.

So I’d like us to take a few minutes of silence for each of you to gather your thoughts about those presentations.

Then, I’d like you to share what takeaways you pulled from your review of those presentations with a couple of your neighbors. What did you see that you’d be interested in incorporating into your presentation next week. Then, as a group, I’d like us to all share as a class.

Sound good?

I’ve also included this example toy presentation by a marketing executive that created this presentation for teaching purposes.

The purpose of this presentation is to secure approval to move forward with an analytics project.

The audience is mixed, and would involve a team of marketers, data analysts, and whomever would be responsible to approve the funding. I’ve invite you to think about how the presentation is structured and designed.

Think about what you can learn from it. And think about how you’d need to adjust the ideas for your own audience and purpose.

Now, i’d like to discuss a few more ideas specific to presentations. Let’s turn to those now.

First, always be aware of what you are asking your audience to read, when you are asking them to read, and do not set your presentations up for them to have to choose between reading and listening.

It’s a common issue.

How have I tried to help you with this issue in my presentations throughout the semester?

In the presentations you’ve reviewed, how did you find the balance between visual text and spoken words? How did they relate?

Now this brings me to a second idea I’d like to discuss. Transitions.

What did I just do, or say, rather, that relates to this slide and when did I say it?

I verbally introduced the next topic before I asked you to see a visual of that topic. The idea is very important and useful, and is formalized in the filmmaking industry. When jumping from scene to scene, it is very effective to introduce that material in some way through sound before sight. That’s called a sound bridge.

Along with verbal introductions, we can provide visual cues, right? In Jonathan’s textbook, he calls this visual material scaffolding. It’s the short, well formatted text of your titles, section slides, and so forth.

There’s a difference between a takeaway document and a well presented slide. The presented slide would be difficult to understand out of context of the presentation. That’s because it doesn’t stand alone. Part of the content is you, the presenter.

If the slides were designed to stand-alone, there would be far too much text to be effective in an oral presentation.

Let’s address the limitations that Tufte discusses.

First, don’t ask your audience to compare things presented sequentially. Instead, present the things needing to be compared at the same time, adjacent in space.

Second, apply the same principles we’ve considered in data visualization to the entire visual communication. Remove anything distracting to increase and reinforce your message.

Third, consider alternatives to just a presentation, especially if you need the audience to have a certain baseline of information to have the conversation.

If you think about it, that’s what we do in this course. I assign you reading before our class discussions so that we can build off of that common information together. So when you need to have the audience primed, give them material ahead of time and use the discussion to deep dive into what you both know.

Forth, play on the advantages of presentations.

Wait, what? I thought you just said they were limited? What are their advantages?

Pacing. Right? We discussed pacing and animation recently. Presentations put the control of pacing into our hands.

When our data or visual is dense, how can we show compare information on a visual? Do we skimp on the comparisons?

No. Right? That’s where we get meaning. Remember this advice?

Ok, so let’s consider some of the design concepts we’ve used elsewhere, and apply them to presentation visuals. Let’s use our Citi Bike case study.

[SCROLL DOWN]

To prepare you for the discussion, recall that we’ve been analyzing rebalancing, and part of that analysis involved reviewing a specific docking station. 31st and 7th Avenue. This station is near both a major train station and major entertainment venue — madison square garden. Right?

Now, if you recall from homework four, the graphic you made near the end was pretty dense, right? So let’s break this down to make it more understandable.

[SCROLL DOWN]

So we’ll start by just orienting ourselves to the measurements. Again, the data we’ll review relates to the docking station at 31st and 7th avenue. And we only pulled data from September 2019, before the pandemic, right?

So I’m representing time on the x axis.

And I’m representing the count of bikes at a given time on the y axis.

You’ll notice that I’ve shaded gray the weekday time periods.

And I’ve drawn a reference line at 47, which represents the approximate number of bikes and docking spots at that station.

Any questions so far?

Ok, let’s layer in the count of bikes at that station over time.

[SCROLL DOWN]

Here is our count of docked bikes over time. You’ll notice a few things. First, you’ll notice approximately five spikes each work week, except four on the first workweek. Those occur each workweek evening.

Next, you’ll notice some odd or intersting gaps.

[SCROLL DOWN]

We have a few ideas of what these gaps may reflect. It could be that the first gap, shown in pink, reflects inoperable bikes docked at the station. And the gap shown in blue may reflect inoperable docking spots.

Or, these gaps may reflect imperfections in the limited data we used to arrive at these estimates. We need to dig into this further.

But let’s assume that the peaks and troughs extend to the limits. So five times a week all the spots fill up, and this makes sense because it’s right beside the train station where people may leave work to head home.

Questions?

Ok, let’s see how we’ve approached rebalancing here. We’ll layer that into this visual in red.

Now we’ve layered our cumulative interventions on the the graphic. Now what we want to compar is when in time the spikes occur versus when we’re adding or removing bikes.

And the pattern we find is that most of our interventions at this station occur on weekends. So we’re not addressing the daily spike in demand for spots.

And that’s an example of how we can layer in information to pace our audience and keep it in context.

Speaking of audience, let’s introduce a basic tool that will help us refine our communications for audience and purpose.

These are templates called user stories and job stories. They’re a way to brainstorm and anticipate what information our audiences need. Let’s see both now.

I’m showing you a user story on the left. It’s a template you fill in to think about your communications. Let’s read together:

As a [person in a particular role]

I want to [perform an action or find something out]

So I can [achieve my goals of]

You can apply this to your interactive communications or to your upcoming presentations to the CEO.

I suggest you fill in several ideas related to your identified audience and use that to guide how to design your communication.

Now, in some cases, your audience may be mixed. So you can focus on a job story instead of a user story. I’m showing you a template for a job story on the right. Again, let’s read:

When [there’s a particular situation]

I want to [perform an action or find something out]

So I can [achieve my goal of]

These templates are really simple. But formally using them can be very helpful in forcing you to be specific in crafting your communications.

And speaking of crafting, let’s just consider how to get to a draft. We’ve talked about storyboards as a tool. And one important characteristic of storyboards are that they take much less time to create than the final communication. You are not invested in it, but it let’s you consider alternatives quicker.

That’s the main point of prototyping more generally. And you can take multiple approaches to prototyping.

Remember the flip book example from Mike Bostock at the New York Times? How they tried out so many ideas before refining?

You can do this with code sometimes. Your effectiveness with code will depend on your proficiency and speed using the tool.

But it’s not the only approach. Some very talented coders, designers, start with basic pencil and paper.

Here is Nadieh Bremer’s desk showing you how she starts her ideas with pen and paper.

Here’s one of her collaborator’s desk. She, too, starts with manually drawing ideas out.

Here’s my desk. I can code very quickly because I’ve had a lot practice. But I still do this because it helps me think in a different way.

Ok, let’s bring together these last two tools and think about how to get help from colleagues.

I’ll call the idea pairwise prototyping.

It’s a pretty straight forward process, but it helps to sketch out the steps.

[go through the slide]

And we can extend the idea across groups to workshop our team projects.

So let’s split up our remaining time. Spend X minutes in your groups preparing to get feedback from another group.

Then, your group will pair up with someone from another group. Go through this pair-wise prototyping together. Split the time and do it for each other. Finally, come back to your groups and share your new ideas with your team before you forget them.

Sound good?