Veri Görselleştirme Bileşenleri | Veri, Kitle, Amaç #8

So when we're looking

at our data and and looking to

communicate with it

i see there is being three key

areas that we want to

we want to focusing on.

We want to look at the data

and we really need to understand the

data.

We need to consider our purpose.

What are we trying to achieve with this data?

And we need to consider our audience.

Who is it that we're trying to communicate to?

And if we look at data for instance to

start with.

Here we have a chart of the

tv ratings for the Simpsons.

And we can see around 2001

there seems to be a big increase in the

ratings of the

television show.

And that ratings increase

might tell us that the television

show got better.

But we should ask ourselves:

Is that reasonable?

Did this television show suddenly get

that much better? And did everybody

immediately understand that it was that

much better?

That doesn't really make sense.

When we go and we examine the data and

we talk to the experts we find out

what really happened there was

the way Nielsen

takes care of their tv ratings they

changed how they collected the data.

And that change in data collection is what's

responsible

for this jump in ratings,

not the actual content of the television show.

So we should be asking ourselves:

Where did this data come from?

How was this data collected?

What is actually being measured?

And does it make sense?

Is it reasonable

when we're examining our data?

We should also be considering our

audience.

This chart type is called a tailor plot.

And it makes a lot of sense

if you are environmental scientist.

But if you are

even if you're someone like me who looks

at charts every day

all day, i still have no idea how to

interpret this.

And if you're going to present to the

lay audience

This is not an effective tool for

communication.

So you need to be asking questions like:

What is the knowledge and expertise of

my audience?

And how interested and invested are they

in the topic or the data?

How familiar are they withthe data?

How familiar are they with the visuals

that you're using?

And what kind of context will be helpful

to them?

And also you should be,

in a lot of cases,

with business applications,

you should be trying to answer:

What decision are they trying to make?

Which leads us nicely into sort of the

last aspect of our venn diagram:

Purpose.

We want to very much consider

the purpose of our visuals when looking

at them.

Here we have an article where the

visuals seem to

contradict the purpose.

This is a visual about

how there are still considerably

fewer female CEOs

in fortune 500 companies.

And the article is talking about

how the glass ceiling persists.

The number of females is

still much smaller than would be

expected.

However this precipitous rise in the

number of female CTOs that the chart

gives the impression of

seems to contradict that message that

they're trying to get across.

When we consider our purpose

some simple changes to the chart...

This is the exact same data

But we've given some context say there

are fifty percent of people are female

so uh let's let's set the axis at fifty

percent.

And now we can see that it has been

rising but that there is still

a considerable amount of difference

between the number of female CEOs and

male CEOs.

And so we get the context of it but it

doesn't

contradict the purpose or the message of

this article

that was written by the Reuters people.

So when we're considering our purpose,

we should be looking at the different

types of visuals that there are out there.

the Financial Times has put together a

wonderful, what they call the visual

vocabulary

available at ft.com/vocabulary.

It provides a number of

different visual types that you can use

to convey your data and the purposes

that they are good for showing.

So if we're looking to convey

ranking,

we can use a number of charts from this

column here. If we're looking at

displaying the distribution

these chart types might be more

effective for us.

And so i encourage you to explore the

different chart types and make sure

that they match with your visuals.

So just to sum it up,

when we're creating our visuals we're

not looking to

to dress them up. We're looking to

pare them down to that core message,

and really

cut away the excess and highlight the

the important. And with that we create

this sort of diamonds of knowledge

that we really want our audience to

get.

If you want to learn more about how to

present with data

these are two books that really got me

started.

Presentation zen is about

presentation visuals but has

a number, a chapter devoted to

visuals with data.

And then

"Show me the numbers" is a fantastic book

by Stephen Few where he

gets more into the science behind

why some methods work and others

don't and

does a fantastic job of very clearly

laying out how to effectively design

tables and graphs that in life.

We also put together a notes page

that you all can access at this address.

You type that in there it's a

google document. It has links to

a number of different visuals from

this presentation

as well as blog pieces, blog articles

that i've written.

And uh you can use that to sort of

highlight the the items in this

presentation. I've given longer

versions of this presentation and so

there will be some additional

information available to you at those

notes.

And that's that's basically it for

what i have

have to present.