1 of 10

Slide Notes

My name is Larry Cummings and I'm a digital product developer.

That means I work with people to figure out what kind of app they should build to help their organization succeed.

Then I help them build it.

When I say App I don't really mean just a program you would find in a mobile marketplace.

I mean any kind of program that would be available for someone to do something useful.

Data Analysis

Published on Nov 18, 2015

I gave a talk at a hackathon in Phoenix about Data Analysis in the context of application development. This is the slide deck I used.

PRESENTATION OUTLINE

Data Analysis

and what it means to your app
My name is Larry Cummings and I'm a digital product developer.

That means I work with people to figure out what kind of app they should build to help their organization succeed.

Then I help them build it.

When I say App I don't really mean just a program you would find in a mobile marketplace.

I mean any kind of program that would be available for someone to do something useful.

Anybody got a peanut?

  • What's the big idea?
  • Data is only mostly dead
  • Information has to be honest
  • To the pain
  • Not to 50!
I'm a big fan of the Princess Bride.

This is our outline.

What's the big idea?

What does your data have that's worth living for?
Your data means nothing.

Your data is inaccurate.

Your data is poorly organized.

Your data is missing facts and meta data you have to have.

Your data will not save your app.

You data, without a big idea is mostly dead.

Untitled Slide

Data is only mostly dead, and mostly dead means partly alive.

What you have to do is make data knowledge, that's because your application must do something useful if you want people to remember it.

Honesty and Information

You can't always tell when your lying
When you start turning your data into information, you will start to summarize your data. You will loose some detail but you will gain power. Your also going to lose some perspective. You won't know if what you did to get more power will mean your information is dishonest.

The best example of this is statistics.

You really only need about 30 data points to establish a trend line in statistics. The thing is that the trend line well be statically accurate, but very low fidelity.

Statistics doesn't care if it's lying. That's one of the reasons it's so popular.

To the pain

Let's do some examples.

Not to 50!

Go slow. Don't rush.

If you have a big idea that's does something useful, make sure that the data you start paying attention to makes that idea more useful.

Most tools for data analysis are really best at removing data you don't want to pay attention to.

Untitled Slide

Thanks and let me know if I can be of any assistance!

Untitled Slide

Untitled Slide