Problems with percentages in population studies

Local coefficient of determination for log-ratio variables

Post by Dr Chris Lloyd

In conjunction with Vera Pawlowsky-Glahn and Juan José Egozcue, two internationally-renowned researchers in compositional data analysis (CODA), I have just published a paper in Annals of the Association of American Geographers (one of the top ranking geographical journals in the world), which shows why percentages cannot be properly analysed with standard statistical methods. Percentages are common in the spatial sciences, but it is not often realised that there are restrictions on how they can be analysed. An obvious example is a regression of one percentage against another  –  in this case the fitted model may predict values which are smaller than 0 or are larger than 100.

Percentages and proportions are referred to as compositional data and complete compositions typically sum to 100 (the case for percentages) or one (proportions). The AAAG article focuses on population studies and, using the example of religion in Northern Ireland in 2001, it shows how population data can be transformed into log-ratios; these new data can then be analysed using standard statistical approaches. The case study gives insights into how the population of Northern Ireland is distributed by religion and it shows that the most obvious geographical pattern relates to the ratio of Catholics to Protestants, although there are also distinct relationships between Catholics and individual Protestant denominations. Whether you are a Physical or Human Geographer, or indeed from an entirely different discipline, percentages or proportions should be used with caution and I hope that this paper helps some researchers make more informed choices.

A Visitor from Notre Dame

Sophie Higham, an A2-level geography student from Notre Dame Catholic College, recently completed a 6-week placement in the Department under the supervision of Tim Shaw and Andy Plater.  Sophie’s placement was funded by the Nuffield Foundation through a bursary scheme that is open to competitive application each year from talented and enthusiastic school students.   The project work focussed on the zonation of saltmarsh foraminifera (or forams) from sites and the Bay of Cadiz, with the purpose of scoping the potential for using forams for historical sea-level reconstruction.  This project links to current research collaborations with IMEDEA at the University of the Balearic Islands who funded an MSc dissertation project in 2011, completed by Frazer Bird.

Sophie Higham from Notre Dame Catholic College proudly displays her gold award-winning poster. Thanks to Suzanne Yee for her assistance in helping Sophie to design the poster.

Sophie learned how to identify different foram species using low magnification microscopy and developed an understanding of the environmental controls on foram distribution, i.e. intertidal exposure, salinity, grain size etc.  The results revealed a very clear relationship between foram distribution and saltmarsh surface elevation – which is the essential basis for sea-level reconstruction using a transfer function approach.  This methodology forms the core of Tim’s PhD research on sea-level trends for the Adriatic coast of Croatia, which in turn follows on from Hayley Mills successful PhD on recent sea-level trends for the Mersey Estuary.

Tim Shaw helped Sophie learn how to identify the various species of saltmarsh foraminifera and to process the results.

Sophie joins a successful cohort of Nuffield Scholars who have been hosted in Geography: Dan Marks (2007), Matthew Sweeney and Matthew Thomas (2008), Stoffer Bruun and Lizzy Goodger (2009).  Their projects have all focussed on coastal evolution and environmental change.  This year, at the Celebration Event Hosted in the World Museum at Liverpool, Sophie received a gold award for her poster describing the outcomes of her research.  Quite an achievement, and very much deserved!

Year 1 Geology and Physical Geographers on a weekend in Snowdonia 2012

Over the weekend of 20-21st October Year 1 students from the School of Environmental Sciences set off for some autumn fun and relaxation in the mountains of Snowdonia, a parallel trip to the Trawsfynedd weekend taken by Geography, Ocean Science and Ecology students. A happy bunch of Year 1 Geology and Physical Geography (GPG) students, along with fellow year 1 students on other Geology and Geophysics

This slideshow requires JavaScript.

degrees and ~9 lecturers to spent two days gallivanting around Cwm Idwal and for some another large hole in the ground (quarry).  Cwm Idwal, a large amphitheatre calved by erosion by ice during the repeated glaciations of the last 2 million years was at its stunning best in the autumn sunshine.

On Saturday after an early start from Liverpool we all congregated at the Llyn Ogwen car park at the foot of the Carneddau and Glyderau mountain ranges. From there Pete Kokelaar led a magical mystery tour through the turbulent volcanic history displayed in the rock record. The geology shows the deposition of huge pyroclastic flows from volcanic eruptions into a marine basin ~450 million years ago, and these strata have been deformed into a large syncline in subsequent mountain building. Outcrop after outcrop were crawled over from the head of the Nant Ffrancon to the foot of the Idwal Slabs. Overnighting in Caernarvon with a good meal, some pretty good beers, vividly colored and tasting shots courtesy of the students (thanks I think…), views of many members of the local constabulary and UK Borders Agency, and some bizarre speckley green-red glitter-ball lighting effects in the chosen hostelry playing havoc with Alan Boyle’s attention span later, a good night’s rest was had by some……

On Sunday the GPG students gained their first immersion into the wonderful world of glacial geomorphology and coring of lake sediments to reconstruct past environments with Rich Chiverrell and Jim Marshall. After a quick introduction to the broad landscape components, the skills of triangulation and geomorphological mapping were introduced, before 2-3 hours of mapping the retreat moraines of the last glaciation to have affect Cwm Idwal 12,600-11,500 years ago. The afternoon saw a switch of focus to the ‘very wet’ marsh surrounding the lake, where a sediment sampler was used to recover ~4 metres of lake deposits. These muds for the upper layers comprise peat and organic lake mud, but quickly give way to blue-grey gritty silts lain down as this last glacier declined and vanished 11,500 years ago.

Physical Geography and Geology interwoven and combined with fantastic weather, great views and some of the finest scenery in the UK; is there a better way to start your degree?

Faerie stories from the Lake District: the MSc field week October 2012

This slideshow requires JavaScript.

The cast list

Prince Charming, Pinocchio, Lost Boy 1 – Richard Chiverrell

Alice in Wonderland, Lost Boy 2 – Dan Schillereff

The Pied Piper, Lost Boy 3 – Neil Macdonald

Overdramatic Damsel in Distress- Laura Crossley

Other Damsels in Distress- Huan Chang and Mary Oni

Sleeping Beauty- Amber Lewis-Bolton

Group 1 – Mary Oni, Kirsten Booth, Richard Walton, Ann Connor

Group 2 – Amber Lewis-Bolton, Matt Williams

Group 3 – Huan Chang, Laura Crossley, Danielle Alderson

Our story begins….. once upon a time (last week), in a land not so far away (the Lake District), the 2012-2013 Masters class in the School of Environmental Sciences set off in a convoy for the annual field class.

For those not inducted into the magical world of physical geography, a true physical geography field trip involves mud… and lots of it! The first day at Hawes Water did not disappoint with ‘competitive coring’ at the margins of the lake. Unfortunately, as we discovered, a competition cannot occur if one team loses their most valuable piece of equipment (a Russian corer stuck 1.3m below the water surface). A valiant rescue attempt was made by ‘Prince Charming’, but regrettably it was not meant to be. Although this was not exactly the ideal start to the field trip, this did provide the moment of the week in our opinion, after he stripped to the waist and submerged into the murky depths to rescue his beloved corer. Evidence is available; but, if required, it is pay-per-view, as poor masters students have to earn a (dis)honest crust somehow. Trust us, it is truly worth it!

Day 2 involved a trip to Roudsea Wood and the coring of an ombrotrophic peat bog. Key discoveries included how to meticulously classify sections of peat cores, and the pattern of changes in bog surface wetness and climate over the past 3000 years. Also important was the parallel invention of a game based on the US version of the sport ‘rounders’, this used corer rods and balls of intertidal clay and is perhaps one for future students to develop. A full scale mud fight was avoided, despite subtle prompting.

Biggar Marsh was the destination for day 3; a salt marsh near Barrow on Walney Island. The aim was use short cores and the record of radionuclide deposition across the salt marsh to discover the controls over the spatial pattern of sediment deposition. A visit from law enforcement officers and the National Coastguard proved the highlight of the day, after local residents believed children may have wandered into the tidal marsh – we were armed with differential kinetic GPS and gamma spectrometers so clearly youngsters in Barrow have expensive toys! The ‘Prince’ managed to charm his way out of any altercation with the law. After a short field day, magnetic susceptibility, XRF and radionuclide measurements were completed back at Castle Head in our makeshift laboratory through which we discovered evidence of peak radionuclide discharge from Sellafield.

Day 4 involved work at Brotherswater, a beautiful lake patrolled by some rather sinister and hungry swans. We used Russian and Gravity corers to take sediment cores from the lake bed from our leader’s vessel. This heavy work was a team effort as the water was 15m deep. We recovered 3m of brown laminated mud and two gravity cores. The now tired motley crew were able to take a well earned break, whilst the ‘The Lost Boys’ abandoned them to recover materials from sediment traps. Three ‘damsels in distress’ were stranded on the vessel as the circling swans moved closer and closer and the others marooned in a slightly safer location at the mouth of the raging Dovedale Beck. Further games ensued, ‘underwater golf’ a game played with a river, a metal pole and a golf ball*. Eventually we were rescued and further cores were retrieved from the lake bed, and treasures of the day were conveyed at a sedate pace to Castle Head for the application of a wide range of techniques (environmental magnetism, geochemistry and pollen analysis) which allowed us to open a window to view the environmental processes of the past 500 years.

Day 5 involved further work at Brotherswater where we recovered more sediment and followed ‘The Pied Piper’ to an eerie, probably haunted, abandoned mine to look for further treasure: lead, silver and zinc. Exhausted and overflowing with knowledge about monks, farming and tree felling. There was an obligatory visit to yonder hostelry, the excellent Brotherswater Inn, to warm up, drink hot chocolate, eat cake, and mentally prepare for Neil’s driving over Kirkstone Pass to get back to the field centre. Possibly the best moment of the Brotherswater days was the look on Dan Schillereff’s face when the cores were unveiled; Alice in Wonderland comes to mind.

We loved Brotherswater, and all decided to return for days 6-7, the group projects. The catchment has a history of lead mining extending back 300 years and the three groups explored the evidence for this and other human impacts on the landscape in various sedimentary records. The first group used a portable XRF gun in the catchment and recorded the geochemistry of surface soils progressing 2km down the valley across former mining areas, current and former river beds, floodplains and agricultural fields towards the lake. The second with some trepidation entered the flooded impenetrable willow and reed swamps fringing the lake delta to explore how the river inflow had behaved over thousands of years. The third braved the vessel once again to take more cores from an alternative backwater location to explore any differences in sedimentation rate. Many hours of laboratory analysis followed and after processing the results, Friday evening culminating with group presentations of our discoveries. The education tables were then turned over as Neil Macdonald gained an education in recent music. Saturday concluded our trip and after extracting a slightly bedraggled ‘Sleeping Beauty’ out of bed, we began our trip on the long and winding road back to Liverpool.

We had a lot of fun on this trip, but we also worked very hard – field work all day and laboratory analysis during the evenings. Despite ‘Pinocchio’ telling us that we would have an early finish every night, this never actually happened, apparently that depends on your definition of early and time is relative anyway. We experienced a range of new techniques on this field trip which will support us in our further study. And as for the ‘happily ever after’ ending, that instalment we must leave for dissertation time in the summer.

THE END…. (or is it).

* Heath and safety: with underwater golf only the ball is submerged not the players, unless appropriate breathing apparatus is available. Life jackets should be worn.

ESRC Funding Success- A Spatial Microsimulation Model of Co-Morbidity

Health care policy

Dr. Karyn Morrissey of The Department of Geography and Planning, in conjunction with the East Kent Hospital Trust, has just been awarded £200,000 from the Economic and Social Research Council (ESRC) Secondary Data Analysis Initiative, Award to to develop a spatial microsimulation of co-morbidity for the East Kent region.

Informed health care policy requires spatially detailed data on health outcomes and patterns of health service utilisation. Health differentials across space arise due to the clustering of individuals who share similar demographic profiles and lifestyle choices into areas where similar contextual characteristics come into play, for example, area-based effects related to low levels of health service provision. To establish the key determinants of health status at both the individual and small area level a large variety of spatially referenced demographic, socio-economic and health data is required at the individual level. Although there are a number of national datasets that contain detailed health and health usage data, these data tend to be aspatial or at too high a spatial resolution to permit health data analysis at the required local level. One solution to this problem that has emerged in the last two decades is the growing use of spatial microsimulation models to represent and simulate health processes at the individual and small area level. Drawing on the experience my experience developing health microsimulation models and with the help of Paul Williamson and Alex Singleton this project seeks to develop a spatial microsimulation for England to meet the information needs of health policymakers and practitioners. Collaboration with the Director of Information in East Kent Hospital Trust has identified the increased rates of patients presenting with two or more diseases (co-morbidity) as a pressing issue, of direct policy relevance, that requires detailed spatial analysis. Co-morbidity is associated with a significant decline in both life expectancy and quality of life and increased costs in the provisioning of health services for individuals. The World Health Organisation has identified increased rates of co-morbidity as a major challenge for health policy. Utilising data holdings made available to academics via the ESRC-supplied data services, this project will develop a spatial microsimulation model of co-morbidity for England. The project, In collaboration with EKHT, will specifically examine rates  of co-morbidity for the East Kent region.

Summer placement at United Utilities: gaining invaluable work experience

Post by Lucy Cheng – Year 3 BSc Geography Student

During the second year of my degree we were advised by the ‘Careers and Employability Service’ to gain work experience and apply for summer placements, as this would stand out on CVs when applying for jobs or graduate schemes.  Although I have maintained a part time job since the age of sixteen I felt that a summer placement would build and develop valuable skills in my desirable career path. However, I understood that the positions where extremely competitive. Several unsuccessful applications later and before I knew it, the second year of my degree was drawing to an end, final exams were finished and libraries returned to peace and calm.  I contacted my dissertation supervisor who gave me guidance on where to look and apply for the experience I wanted.  This proved that you don’t get anywhere without asking! My supervisor pointed me in the right direction which led to me obtaining a summer placement with United Utilities who provide around 7 million people and 200,000 businesses with water and sewage services in the North West.

The first week of my placement was a shock to the system, a full time job, Monday to Friday, in a new working environment.  I was assigned to work in the ‘Drinking Water Regulations and Public Health Team’ who gave me an extremely warm welcome and made me feel very comfortable.  I was assigned my own project to work on for around 10 weeks which involved speaking to people from all sides of the company both on the telephone and face-to-face.  I was very lucky to be given the opportunity to work on a project that allowed me to see all aspects of the company and even visit treatment sites around the North West.  Every day I learnt new skills and built on knowledge of the company, either in courses, meetings or through talking with colleagues and no two days where the same, there is always something new to learn.

During my time in United Utilities I was able to contribute to a team away day volunteering in Quarry Bank Mill where, as a team, we replaced fencing within the National Trust site.  This was a great team-building day outside of the working environment, we did a great job although the next day we were all aches and pains.

During the summer I was also researching for my water quality based dissertation. However, due to the amount of rain that occurred over summer, it was difficult to obtain the data collection necessary to support my research.  With the support of United Utilities and guidance from colleagues in various teams I have been able to alter my dissertation and research a topic to one that is both beneficial to the company and interesting to myself.

United Utilities has offered to keep me on until Christmas working one day a week whilst completing my final year of study.  I believe that the point to take from my summer experience is to never give up on what you would like to achieve and to never be afraid to ask for guidance.  Both United Utilities and the University of Liverpool have been a great support to me in achieving my ideal career.

The Role of the Marine Sector in the Irish National Economy

Post by Karyn Morrissey

The realisation that the world’s oceans play an important role in climate regulation and many territory activities, notably food production, coupled with economic changes and the rapid advancement in ocean technology have seen a shift in the perception of the importance of the marine resource. My recent research in Ireland has estimated both the national and regional economic value of the marine sector. However, economic activity does not exist in a vacuum. Activities in the marine sector not only directly affect the industries in the sector but also influence other sectors through inter-sectoral linkages. My new research uses an Input-Output (IO) methodology to examine the linkages and production effects of the Irish marine sector on the national economy. This work is presented in the world’s leading marine policy journal, Marine Policy, (volume 37, p. 230–238) and represents the first effort to quantify the inter-industry linkage effects, production-inducing effects and employment multipliers in the marine sector. I found that a number of marine sectors, notably the maritime transportation sector, have an important economic role within the wider Irish economy.


Visualisations of Geographical Data On The Web – Is What You See Always What You Get?

Post by Chris Brunsdon

I’ve been involved recently in looking at ways in which geographical data (and particularly Human Geographical data) has been represented on the web.  I think it is important to look at any data analysis or visualisation critically – are there any factors which have been overlooked in the collection of the data used, or in the way it has been represented graphically, or modelled statistically,  that might lead to misinterpretation,  for example.  In the UK and a number of other countries,  there have been recent moves towards the provision of open data – where data collected by governments is made freely available over the internet – here is an example.  Also,  some international organisations provide this data. Many providers of web sites use this data,  linked to maps and other visualisations,  to provide easy to access overviews of patterns in this data.  Here,  I will show two such sites – these are all interesting,  although there are reasons generally why the critical approach I advocated earlier is valuable…

The first is the web site – showing details of recent crime in an area.  I think it is a great advance in making data publicly available,  but one major problem is the way data is represented if you ‘zoom in’ on the map.  Here is a  map showing crimes near to Liverpool Geography. If you zoom in to the area very close to the marker balloon, it shows the exact location in streets where offences have been committed – for example 3 in the middle of Back Mulberry Street,  if you are looking at the map for August 2012.  But does it really show this?

Quite likely not.  A small caption underneath the map reads “Crime and outcomes are mapped to an anonymous point on or near the street or location where they occurred.”  In fact,  to preserve some degree of anonymity small groups of houses or other buildings are sometimes ‘clumped together’ and a point chosen in the middle of the clump.  This is quite likely what happened  in the  example above.  This can make it appear as though a mini crime wave is occurring in front of a particular house,  or on a street corner,  when this may well not be the case.  This isn’t the first time the issue has been raised.  I’m not sure that a small disclaimer underneath the map entirely compensates for the striking but ultimately misleading visual impact of the map.

The second example is less problematic – see here.  This web site gives population pyramids for countries across the world at regular time intervals from 1950 to 2100. It is quite illuminating to compare, say, current pyramids from the UK and Nigeria.  If you are not familiar with population pyramids as a model of population structure,  the site provides a link explaining this,  and suggesting interpretations for common shapes.  The site also shows total population growth from 1950 to 2100 in graph form,  as well as letting you select pyramids from different points in time.  It is also interesting,  for example, to compare the predicted population structures of Nigeria and the UK in 2050.  Perhaps my one issue with this site is the ease from which you move from actual past figures to predicted future figures – since the latter are based on models they will be subject to a fair amount of uncertainty.  Perhaps the graphics could give some kind of visual clue about this.

In both cases,  I think it is excellent that the information has been released – and the organisations responsible for this deserve praise.  However,  I would always advocate looking at any graphic with a critical eye.

Latest QWeCI Project Newsletter now available

Post by Andrew McCaldon

I am the project secretary and Dr. Andy Morse is the coordinator of the QWeCI Project – Quantifying Weather and Climate Impacts on Health in Developing Countries.

In this project, researchers across 13 European and African research institutions work together to integrate data from climate modelling and disease forecasting systems to predict the likelihood of an epidemic up to six months in advance.  The research, funded by the European Commission Seventh Framework programme, focuses on climate and disease in Senegal, Ghana and Malawi and aims to give decision–makers the necessary time to deploy intervention methods to help prevent large scale spread of diseases such as Rift Valley Fever and malaria.

Read about the recent activity in the latest QweCI Project newsletter, which can be downloaded here, and more information can be found here.