All language subtitles for 04_graduated-symbol-maps.en

Graduated symbols are similar to proportional symbols,

but they are using classes.

In other words, instead of scaling

each symbol proportionately to the size of the value for that dataset,

it's first grouping those values into classes and then

assigning one symbol size to that class,

so there graduated in that way.

I think the easiest way to see that is to look at

the legend and see that these are now classes.

So, in other words,

this symbol size represents any value that's in that range as

opposed to the proportional symbol legend where they are examples of sizes.

So here we have three different sizes for our classes,

and there are only three different sizes of symbol on the map,

and that's it okay.

So, there's good things and bad things about this.

One of the things is that that's an advantage is that we don't have to worry about

appearance compensation because people are not

going to be trying to estimate specific values from these symbols.

All they have to do is be able to slot that symbol into

a category or a class to say it's either in this case small medium or large,

even if there was like five class sizes.

As long as they can match the symbol to the legend size then

they'll be able to estimate what that value is within that class,

and so you don't have to worry about appearance compensation.

The other thing that's I think useful is that it's

much easier for someone to interpret patterns in the data.

So, all they're really looking at for example here a small medium or large that's it.

They don't need to necessarily look at

all the little subtle nuances or this one's slightly bigger than that one,

so it's a good news bad news thing.

Some people really don't like graduated symbols because

for example if you have a symbol that's twice the size of another symbol,

it's easy for people to assume that means that the value is twice as much.

With graduated symbol that's not necessarily the case,

it's just that's the size of the symbol for that class not for that particular value.

So, some people like them some people don't,

I think it depends a bit on your dataset,

and your intentions, your audience,

what is it you're trying to get across whether

proportional symbol might be better than graduated symbol.

Either one can work you really I think have to play

with them and get a sense of what works better for one versus the other.

If we compare the same data using proportional symbols versus graduated symbols,

this is the difference in terms of what we see.

So let's have a closer look at this.

With a proportional symbol,

you will notice that Toronto is much larger

than Von which is a much smaller city relatively speaking.

But in the graduated symbol map,

they have ended up falling into the same class.

So according to the graduated symbol map

that's implying that they are essentially the same.

They are in the same class, they have the same size symbol,

therefore, their values must be similar to each other.

But of course, that's not really the case if you look at

the difference in the proportional symbol map, it's pretty dramatic.

This is a case where okay well maybe then we should be using more classes on

our graduated symbol map maybe there's ways to mitigate that affect sure that's possible,

maybe the class boundaries could be a little bit different.

But I am using this here is a good example of saying, well,

anytime you make a map like this you should think

about these things look at the effect that it's had,

what's happened to the data?

What's the effect I'm getting?

I'll show you one other example while I'm at it

is this one here of Waterloo in Kitchener.

So with the proportional symbol map you'll see that Waterloo is smaller than Kitchener,

but not that much smaller.

But here they've fallen into two different categories, two different classes,.

So now Waterloo looks like it's this tiny little hamlet

next to Kitchener which is much larger according to the graduated symbol map.

So, we can end up with

two quite different values in the same class implying that they're the same.

We can have two values that are

a little bit that are similar to each other and going up in two different classes,

so they look more different than they are.

So, you really have to experiment with

this and see what you get as to what you think is going to work

best but I wanted to point that out is that you have

these options in terms of what might work better proportional symbol or graduated symbol.

When symbols go bad, so,

what I'm trying to show here is that if you use proportional symbols,

if you have a value that's really high compared to the other ones,

or if the smallest symbol size that you've chosen is too large,

you can end up with these ridiculously large symbols

which of course is a little too much for a map like

this it's great it's an effect as a little map humor let's say.

But if you're actually making a serious map, this would be no good.

It's not going to help anybody interpret the values between these different ones.

So this would be an example where I graduated symbol might work better if you

do have one or two or few outlying values that are much

higher than the rest of them then you would end up having to do with

graduated symbol is have them in the same class with the same size symbol,

and you won't have the proportional effect of having one, or two,

or three, or whatever of these symbols be ridiculously large compared to the others.

So I just thought I'd point that out.

I also wanted to mention that you don't always have to use circles,

you can use any symbol that's available in ArcGIS.

So here I just for fun,

this is actually I'm going to admit that I'm using the

proportional or the population data for this.

But if you pretended that these were a number of flights per year or something like that,

then you could use airports symbols to relate to number of flights.

So you can have a little fun with this to make it a little

more interesting than just having typical kinds of circles.

There is the effect as I mentioned earlier that with squares people

don't have as much of a compensation effect problem there so,

if you use squares instead of circles hey that's a little side benefit.

Just for kicks as the last one,

I thought I would use something that kind of looks like three-dimensional spheres.

Here I've I'm charting the sales

for my marble corporation that I have just having some fun.

So, again especially with a nice simple light background like this,

it makes the spheres pop out a little bit more.

Depending on the audience,

and the intention of the map and what it's for you

may want to experiment with something that's a little more creative like this.

So some potential problems with symbol maps proportional symbol maps,

if you have too many values,

it can be hard to tell them apart.

So if you have hundreds or thousands of values

proportional symbol map may not be the way to go because it can be a bit overwhelming.

If you have high values as I was mentioning that can obscure some of the other values,

so that's not going to work very well.

Probably one of the worst things is if your data all looks the same,

or if you're symbols all look the same because

your data values are all similar to one another,

then it is going to look like a really boring map, and it's monotonous.

I suppose if that's the message you're trying to get

across if you want to just tell people look they're all similar to one another,

you could do that but then I would say well do you really need a map

to say that maybe you could just say that in

one sentence in a report or something like that.

So, you have to look out for these things is you want to

have a pleasing amount of variation,

you want to have a good range of sizes,

you don't want it all looking the same,

you don't want anything too big or too small.

If you can do that in a Goldilocks happy medium way,

then you can end up with a really nice looking map.