Where does outreach funding go?

Over the past 17 years I've been based in Manchester and Cardiff. During my time in those places I was involved with a few public engagement grants from the Science and Technology Facilities Council (STFC). These grants are called "Small Awards" (there has also been a "Large Awards Scheme" in the past) and are open to anyone who wants to promote the science areas covered by STFC. They are awarded in two rounds each year. For instance, I was on a team that got three STFC Small Awards to create and improve the Jodcast.

I've benefited from the STFC Small Awards but something has niggled at me over the years. After each round is complete I look at the list of funded projects and I've noticed that very few were from the area I grew up in - Yorkshire. I've been meaning to check if this was just my perception or was actually a thing and I've finally gotten around to it. I have converted all the PDFs listing previous award winners between 1999-2014 into a simple data table, written some code, and started to visualise the data. Here is a heat map of UK funding scaled by the population in each region.

UK map of STFC funding
Map showing STFC Small Award funding between 1999-2014 scaled by population in each UK region. The scaling is linear with the North East at 0.09 pence/person/year and Wales on 0.56 pence/person/year. For the numbers see the table.


These data show that the North East and Yorkshire don't get an equal share of physics outreach funding. It is a little surprising that Leeds - the UK's 3rd largest city - only got 2 outreach grants over a 15 year period. Those were both for a theatre company that did national tours - none involved the university. So is this a bias in STFC? Given the make-up of the Small Award panel I really don't think so. I suspect that the bias is at the application stage with few people from the North East and Yorkshire applying.

If you are from Yorkshire or the North East and want to communicate physics, please consider applying for a Small Award. The next round closes on 9th October 2014 at 4pm BST. Go apply.

Posted in astro blog by Stuart on Wednesday 27th Aug 2014 (16:40 BST) | Add a comment | Permalink

Comparing areas

On Twitter over the weekend I saw a couple of retweets of an infographic comparing the money raised to combat various diseases and the number of deaths caused by those diseases in the US. The infographic is interesting. It is also unintentionally misleading within each column.

Money raised vs deaths
An infographic comparing money raised for various diseases and the deaths for the same diseases. The original source is unknown. I have included the image here for critique.


The problem stems from the fact that 2D shapes (circles) have been used to represent the values but the infographic artist has scaled the shapes by their diameters. When we look at the graphic we naturally compare the areas of the circles. Unfortunately, areas scale as the square of the diameter: if a value is twice as big and you double the diameter, you've made the area four times as big. As a result, this infographic distorts the relative values. For instance it makes it look as though breast cancer accounts for about 72% of the money raised when it is actually about 51%. In the deaths column, it looks as though heart disease accounts for about 92% of deaths whereas it is actually 64%.

I've made a version of the infographic to show the difference between scaling the circles by diameter and scaling them by area.

Money raised vs deaths
A comparison of comparisons. The money raised and the deaths with the circles scaled by diameter (misleading) and by area.

The moral of this post: if you are using 2D shapes to show relative values, make sure you scale them correctly.

Posted in astro blog by Stuart on Sunday 24th Aug 2014 (23:35 BST) | Add a comment | Permalink

Expertise

The UK's Daily Mail reporting is shoddy at best and their content can be malicious. They make basic factual errors and aren't keen to correct mistakes. That is probably a given and I'm pretty used to the baseline level of distaste I have towards them. However, today I got much more annoyed that usual when they heavily implied that two astronomers were invited onto the BBC's Newsnight (a 10.30pm news programme) to talk about the recent BICEP2 cosmology results because of their gender and/or nationality.

The Daily Mail opinion piece specifically mentions the genders of Dr Maggie Aderin-Pocock and Dr Hiranya Peiris. It contrasts them to the genders and nationalities that the writer assumed for the BICEP2 team. Why do this? What relevance do the gender and nationality (or race) have to do with their ability to talk on the subject? The Daily Mail piece doesn't mention the gender or nationalities of the three other astronomers that were on the Newsnight BICEP2 item. The expertise or presence of those astronomers wasn't questioned based on their genders or nationalities. In fact, they weren't mentioned in the Daily Mail piece at all. By omitting to mention the three male contributors, it left readers with the impression that men were somehow discriminated against. This fits the distorted reality of the Daily Mail (and a false but popular narrative that men are being oppressed) who clearly want to make a political point against the Newsnight Editor. The implication that the women were only there because they were women is demonstrably false (see below). The implication that only women were on this piece, is also false.

Let's get some things straight. It is OK to ask if people are qualified to talk about a complex topic. It is not OK to only challenge the scientific qualifications of people of one gender, of one race, or any other irrelevant physical characteristic. That is discrimination however you dress it up.

For the record, Dr Aderin-Pocock is an astronomer and space technology engineer who has worked on Gemini/JWST and co-presents the BBC's Sky At Night. Dr Peiris is a cosmologist who has worked on the two big cosmology space missions of this millennium; WMAP and Planck. She has worked on Herschel-SPIRE. She is a provisional member of the Dark Energy Survey. She is PI on large research grants. She has had a Hubble Postdoctoral Fellowship, an STFC Advanced Fellowship, and won a bunch of prizes. If that doesn't make her qualified to comment about her own area of expertise on a general news program, what does exactly? Both are more than qualified to be on Newsnight talking about astronomy/cosmology.

There is an excellent open letter from the Vice-Provost for Research at UCL to the Editor of the Daily Mail, Paul Dacre which I saw via a rightly outraged Chris Lintott. Kudos to UCL for defending their research staff against discriminatory pot shots from newspaper commentators.

I'd write to the Press Complaint's Commission but it looks as though the Editor of the Daily Mail is the Chair of the PCC Code Committee so the chances of him enforcing something against himself seem a little slim.

Update 2014/03/21: How wrong could one short article be? As well as the things I mention above, I should have also taken issue with the Daily Mail claiming that the BICEP2 team were "(white, male) American". Here is a picture of some of the BICEP2 team. They are clearly not all men and white. I even know two of the team based at Cardiff University (in Wales). Finally, the Daily Mail actually quoted Dr Peiris in an article a few days before but "Ephraim Hardcastle" (alter ego of Peter MacKay) didn't use that to claim bias at the Daily Mail. So the evidence points to lies and racist statements for political point scoring.

Posted in astro blog by Stuart on Thursday 20th Mar 2014 (16:40 GMT) | Add a comment | Permalink

Multi-wavelength Universe Activity

When you look up at objects in the night sky you only see part of the picture; you see visible light. In reality, the universe is emitting light at a huge range of wavelengths from radio waves to gamma rays and everything in between. It is important for astronomers to compare the light of each type from an object such as a star or a galaxy as that tells us huge amounts about their composition and conditions.

To help students explore the multi-wavelength nature of astronomical objects, the UK Herschel website has a multi-wavelength universe activity with a bunch of educational resources for teachers. The activity was originally written in Flash and Chris North recently asked me to convert it to Javascript/HTML to make it easier to update and so that it would work on devices that don't support Flash (e.g. iPads/iPhones). In refreshing it I also made it responsive (it should reformat itself for narrow devices such as smart phones) and built in support for different languages (you'll be able to use it in Welsh at some point soon).

We made the whole thing open source and, if you've checked out a copy from Github, you can even run in a browser locally without needing an internet connection.

Let us know what you think. Please report bugs via Github or tell me on Twitter.

Posted in astro blog by Stuart on Wednesday 12th Mar 2014 (13:22 GMT) | Add a comment | Permalink

APOD Linked Data

This is a very esoteric post about Astronomy Picture of the Day (APOD) and data formats so if neither of those interest you feel free to look at a pretty picture of a nebula instead.

Since 2007 I've been running the @apod Twitter account. Since the US government shutdown of October 2013 I've been providing an APOD search engine too. As a result I have put some effort into parsing the content of every APOD since 1995 and have them in a big file split into fields such as date, title, author, description, URL and astronomical objects (coordinates/type). I realised that this big file of data could be useful to other people so was trying to think of a good way to share it.

A few weeks ago I stumbled across JSON-LD which stores text in Javascript object notation but with a few twiddles to describe what the data actually are and help link data from different sources. I'm familiar with JSON and the approach of the JSON-LD folks seemed easier to grasp and work with than the alternatives from the RDF/SPARQL/Linked Data world.

After a bit of playing I had converted my file into a first attempt at JSON-LD. I uploaded it to GitHub and mentioned I'd done so on Twitter. A little while later I was pleasantly surprised to have a tweet (out-of-the-blue) from Markus Lanthaler - one of the authors of JSON-LD - suggesting some improvements to what I'd done. He sent me a pull request and I updated things accordingly. It was brilliant to receive helpful input like that. Huge thanks to Markus for being so proactive and helpful.

If you play with the APOD JSON-LD file let me know what you think or if you have suggestions for improvements.

Tags: | |
Posted in astro blog by Stuart on Friday 14th Feb 2014 (17:04 GMT) | Add a comment | Permalink