Pie charts!

HE Festival of Data 2019: Part 3

It’s not often that you get the feeling that you are in a room with a couple of hundred like-minded individuals. This happened at the HE Festival of Data at the University of Huddersfield during a talk by Jackie Njoroge, Director of Strategy at the University of Salford. During Jackie’s talk on the need for timely data to ensure strategic delivery she said in passing that “anyone who presents anything in a pie chart should be …”. This statement met with a round of spontaneous applause, in which I joined, which will course of course come as no surprise to anyone who has heard me lecture on the misuse and misinterpretation of data presentation. My own attitude towards pie charts was crystallised by my PhD supervisor who would always cross out a pie chart in a draft whenever he saw one. Just for fun, and to make the point that they do have valid uses, I included 3 pie charts in my final submitted thesis. This does not detract from the fact that the potential for mis-interpretation with pie charts tends to bring on a rant in a lot of data-oriented people such as myself. Why is this? Well, amongst other things we, people that is, are not very good at judging and comparing angles. Stephen Few has written an excellent piece on pie charts with more detail on this. 

Interestingly we then learned later in the day about another of the issues with pie charts and that is their popularity. People like them, or more specifically, students like them. During their workshop on student dashboards Liz Bennett and Sue Folley from the University of Huddersfield walked us through some work they did with students on the content and presentation of data for student dashboards. Some of this work was published here. One of the findings of their work was how popular pie charts are among students for the presentation of information about them. So we find ourselves in a dilemma. Pie charts are popular but can be mis-interpreted. Should we use them or not? Well, it seems that if the goal is to engage students in their own data then we should use data visualisations they like to look at so pie charts may need to be included. We can always present them electronically which could allow us to add numbers to the charts with, for example, text that appears when you hover over them, reducing some of their drawbacks. It is always important to remember that the goals of data visualisation include not only the need to make it as easy as possible to understand the message in the data but also to engage the viewer in the data. Despite how we may feel about them it looks like pie charts are here to stay. 

The things we have to do to get our message across.

Educated but Ignorant

HE Festival of Data 2019: Part 2

“Hello, I’m Dr Tobe and I’m educated but ignorant”.

I think this will have to be my new opening line in conversations. A great way for people in Universities to describe themselves. But where did I get it from? Well, I’ve just been to the HE Festival of Data at the University of Huddersfield and found it to be a fascinating and enlightening event. The phrase comes from one of the main talks, this one by Sally Turnbull, Director of Planning & Insight at University of Central Lancashire (I’ll be writing about some of the others later). Sally was talking about University League tables. I suspect a lot of people in Universities would consider that to be a dry as dust subject and would rather gnaw off their own leg than attend a talk on such as subject but this one was well worth the effort. Sally had been tasked with writing ‘A Guide to UK League Tables in Higher Education‘ and when she asked about the target audience was told it should be aimed at the ‘educated but ignorant’. Most people reading this would probably consider themselves to be educated, but how many would admit to ignorance? Yet as Confucius once said, (yes really, he did), “real knowledge is to know the extent of one’s ignorance”. Where the creation of University League Tables is concerned I am prepared to admit to the astonishing depth of my ignorance. Thankfully Sally, through her talk and report, has done something to knock the edge off that. 

This isn’t the place to talk about all the data in the UK University League Tables, you can read Sally’s report for that, but I thought I would share a couple of tidbits from her talk. 

‘Completion’ is included as a new metric in The Guardian League Table, so is not in the report. It is not, as you might imagine, the number of students who succeed in their studies. It is, in reality a modelled projection of what could be happening about now, based on data from a few years ago. Wait, what? Is that going to be even close to reality? What if things have changed?

‘Student:Staff Ratios’ are important statistics but have a number of issues in the league tables. Perhaps the most troubling is that the only ‘coding system’ available to link staff to students is the HESA cost centre coding. A significant problem with using cost centres is that HE Institutions may not code staff and students to the same cost centre. Hmm.

Tales like these about the data shouldn’t distract from the fact that producing the league tables at all is a very difficult undertaking. They are trying to simplify an extraordinarily diverse sector and they do tell us something worthwhile about what is going on in Universities, just not necessarily what people might think they are saying. They also don’t really tell prospective students and their families what they want to know. So what do prospective students really want to know? According to Sally they want to know:

  • What do they need to do to get in?
  • How big are their classes going to be?
  • What are their fellow students going to be like?
  • What do others do with that degree after they graduate?

Let’s see some of that in league tables.

Inspiration

HE Festival of Data 2019: Part 1

Do you suffer from post-conference time distortion?

You have just been to an excellent conference and you come away bursting with ideas and inspiration, you’ve networked well with people who are interested in the same things as you and then you return to work… Some months of distraction later you look back and think: what happened there? Welcome to post-conference time distortion. This has happened too often to me, with a few notable exceptions.  I am determined that this time things will be  different.

I am on my way home from the HE Festival of Data at the University of Huddersfield. There were some excellent talks and workshops and I thought to myself ‘I must share these with my colleagues’. So returning on the train, leaving aside the fact that someone apparently shooting at the train had delayed it a bit, I began to write up my notes. As I looked at the profiles of the speakers I began to read some of their work and importantly their blogs and realised that here was the perfect way to report back. I’ll start blogging about it. So I started to write and here we are.

So who were these speakers who’s talks will be inspiring a series of blog posts here? 

  • Mario Ferelli (former Director of Analytical Services at HEFCE and former Operational Research Consultant at the Office for Students) on “How to survive the data jungle”,
  • Andy Youell(former Director of Data Policy and Governance at HESA) covered the sector’s relationship with data,
  • David Kernohan(Associate Editor of Wonkhe) looked at the effect of data on student choice,
  • Jackie Njoroge(Deputy Chair of HESPA and Director of Strategy at University of Salford) emphasised the need for timely data and
  • Sally Turnbull(Director of Planning and Insight at UCLan) gave us an insight into UK League Table data.

I hope you will come back to hear more about what they said.