Now there are different types of graphs but overall what are good graphs?
How do we produce good
graphs? What are the things that we need to keep in mind? Again it's an art form, it's not an exact science but good graphs are always clear and well labelled. Think about do I have titles on my graph? Do I have access titles on my graph? Do I need to have those additional labels like on top of the bars like those data point labels? Maybe, maybe not, but think about that. Good graphs are concise and informative. Think about why are you doing the graph, what is the purpose of the graph, and based on that try and focus the reader on the substance, what you want them to extract from it, not the form not the pretty colour or the whatever it is. Now we use colour, we use nice pictures, we use all of that to help make it more attractive but fundamentally it's about the substance. What is the information you're trying to convey? Then, we use nice colours and help it out but it's not just all about fancy nice colours that look good but don't display the information, that's not okay. So step one, we want to be concise, informative, focus on the substance then we make it look pretty. Now generally good graphs encourage comparisons between categories, trends over time, usually that's what we're we're looking at because visually our eyes are very good at picking comparisons. And finally good graphs should be accurate, they should not be misleading, we should use accurate figures, and when we talk about ethics we'll talk a little bit more about that. So, graphical techniques, there's a saying in English that a picture is worth a thousand words and that's very true here. So we're trying to summarise a lot of information, it would have taken tables and lots and lots of text to convey this but we show one graph and the reader can quickly see all that information. Now as managers you’re time poor, so this is really helpful to you to find out that information quickly and efficiently to communicate it with your team, with your bosses, with other managers. So really think about and when you're writing reports, whatever you're doing, is there a graph here when you're presenting data, is there a graph that might be a nice way to present this that will quickly and easily communicate this information. When you're asking people to prepare and do some data analysis for you, yeah tell them to think about hey I would like to see these results in a graph because that's easier for me to quickly see it. Then give me all the other information and the backup information but a nice summary graph can do a lot of good. But we need to use the right graph. So we've seen bar and pie charts, really good when we can break things down into categories and either focus on the contribution to the whole or comparing the different categories. We looked at histogram, for quantitative variables, numerical variables, the numbers breaking them into categories then we can look at frequencies and relative frequencies. We looked at line plots for a time series graph so we could see what was happening over time. Finally, we looked at a scatter plot to see the relationship between two graphs. But now it's time to actually do this in Excel. So now that we've seen these let's have a play in Excel, to make sure we can calculate and produce them in Excel. Good news is, very easy, so not too difficult but there's lots of different options so you need to spend some time in playing around to see what's happening.