AutomaTable? Smarter tables that are easier to visualise and automate

Paper tables - not automatable

Before Christmas I spent some time working with tables on a seven-country study, and going Grrr. It seems to me that the main purpose of tables is to provide the aggregated data that populates a visual report. I’d be interested to hear from anyone who disagrees, but in my experience the research results are really only analysed once charted, making the tables a passing phase on the reporting process flow and a repository for looking things up rather than the tool which is used for interpreting the data. So, why not design them so that they are easier to use for reporting? Here are some ideas for doing just that.

First: use the text you want to see in the report, not the text from the questionnaire. Text in tables should be abbreviated and converted into PowerPoint chart-friendly statements, not whole sentences lifted from the script which will never appear in the report itself. Here’s a version from real life of a table title:

“S10. S10. Now, looking at this list, please tell me which of these products/solutions do you know, even if only having heard about them? * QUESTYPE. Questionnaire Type + SEC + S3. S3. In which city do you permanently stay + S2. gender + AGE + Awareness_type + Usership_Type + Usership_Criteria_2 + Usership_Criteria_1 + Stated_Condition”

In the report, you would probably want “Prompted Awareness of Products” with possibly the question identifier. That’s all. That should be the question text in the table as well, so that you can use the text that’s there either by copying and pasting or lifting it via a simple automation program.

Second: apply the same reasoning to column headers. If you want it to say “All respondents” – specify that in the tables, plus the relevant, abbreviated wording for the subgroups you’ll be using. I still see columns where words have been separated oddly with hyphens so that the header text fits nicely over the column. We really need to stop doing that. It’s much more important that the text matches what’s required in the report than looks neat on the page. I am assuming here that people are looking at tables in Excel or a similar copy and pastable-format – not a pdf or a printed page. Does anyone print tables any more?

Third: – and related to above – don’t separate a long banner into shorter banners and have them one below each other on the same sheet. Have all the columns you need in one long banner that stretches off to the right. That way each question is only on the sheet once, which makes it much easier to work with. We don’t have to worry any more about whether it fits onto a page. We can just freeze the first column and scroll across these days, can we not.

Fourth: Keep the number formats consistent. Most research reports contain column percents plus a few mean scores and unweighted base sizes. Remove everything else, particularly absolutes. If you’re working with statistical significance indicators, have them output into a different sheet so you don’t have to remove them manually if you don’t need them. You want to be able to highlight a block, including text and figures, and copy and paste it all as one movement, not constantly backwards and forwards because there are blank rows or absolutes mixed in with percentages. Speaking of which, decide whether you want percentage signs or not, thoughout the end report, and have them in the tables, or not, accordingly. I’m a little fed of up pasting into a blank Excel sheet and doing (=B1*100) 15 times in a day…

Fifth: Give each question a title – a single title, not two parts. That has two advantages – it means that all your tables will exactly match in terms of layout and the column headers will always be in the same position relative to the title text etc. And it means the title will be more useful in the report. In other words, don’t have this:

“Q2X1. Q2. Please tell me how much you agree or disagree with each of the following statements about your attitudes towards life and personal care in general.

I buy environmentally friendly products”

Have this:

“Agree/Disagree: I buy environmentally friendly products”

Sixth: Order scale questions how you will need them in the report. This will need a bit of forethought in terms of the PowerPoint charts you are using. For example you may want to work negative to positive left to right, with DK/NA at the end. Or the other way round. Whichever it is, get the tables set up to match so that you are not constantly repeating the same copy/paste operation.

Seventh: Don’t rank or zero suppress, particularly if you will be working with more than one set of tables across a project. the more similar they are, the better. Besides, you will probably want to rank on different things depending on what you are reporting, not just the total column all the time, so better to do this in the chart datasheets or via automation.

Eighth: If you have an index page at the front with tables on different worksheets, this can help navigation enormously. Make sure the names of the tabs relate to the question and will always match that question even if new questions are added. ie, not just Table1, Table2 etc, with the danger that adding a new table will mean that they no longer match the content of the sheet.

Usable and automatable tables will be a lot easier to work with manually, and also very easy to use in an automation package without a lot of tweaking and moving stuff about, thus allowing more time to actually look at what the results mean rather than spending tedious hours just getting to a stage of having charts.

 

 

 

 

All I want for Christmas is…better PowerPoint

christmas-present better PowerPointIt’s not asking much. I’d just like to be working with a data visualisation package (since that’s what the MR industry uses it for) that’s a bit fitter for purpose. And I do admit that PowerPoint 2013 is better than previous versions, for example on setting page size, and the eye-dropper. However. There are some things that it just can’t do that, given how we use it, it should be able to do. So just in case Microsoft Santa is reading this, on my Christmas list of better PowerPoint would be:

Charts

  1. Stop being such a pain in the neck when I try and set source data. Make it easier to specify what are the axes, what are the series etc etc and don’t reset these things every time the chart data area is changed. And it would be nice to be able to do this without going into the datasheet. Specifying which parts of the chart does what should be a lot easier and more intuitive.
  2. I’d like to be able to specify some of the dimensions within the chart, IN PARTICULAR the width of the vertical axis label field. Fed up of having to shorten the wording so the label on a horizontal bar chart doesn’t wrap. If I want a long label, I should be able to have one, so say I. Without having to resort to creating a table next to the chart for the labels and hiding the actual labels because I can’t get them to do what I want. No more! And I’d like to be able to specify the alignment as well. And on all chart types, even spider charts.
  3. PowerPoint has become much better when it comes to lining up shapes, with its automatically-appearing guidelines and grid options. But charts don’t seem to be able to join in the fun. I’d like to be able to align charts in the same way as other objects, and, if it’s not too much trouble, choose whether I’m aligning the chart area or the plot area. And I want to be able to decide how big the plot area is within the chart area. That way, if I have two charts next to each other that have to be the same size, I’m not completely stuffed if I alter the plot area on one of them, because I have no guaranteed way of ensuring that the plot area of the other one exactly matches. Unless I copy and paste the chart I’ve just changed and start again…
  4. Formatting charts. I wonder how much of my life I’ve spent doing this. How about a chart format paste tool, where all the formatting options of one chart can be transferred onto another? It would include series colour, effects, line colour and style, labels and position, font, axis options etc etc etc. That would be neat. Especially if you could do it across different chart types.
  5. I’m in a phase of wanting to place lots of small charts next to each other on the same slide. It looks great (I think) but is a pain, because each chart datasheet is completely stand-alone. How about an option where a chart datasheet is made available to all charts on a slide? Then you can have all your data in one table and use it to build all your similar charts without having to copy and paste the data n times. There could be other ways of making it easier to work with a number of different charts at the same time. Please, let’s drop this idea that there should be only one chart on a slide.

Tables

  1. PowerPoint tables. Where to begin? How about some table functionality for a start? A simple sum feature, and count, average calculations? Even Word can manage that. Some ranking, maybe? Would be nice.
  2. Copying and pasting into PowerPoint tables always seems to create work. It doesn’t seem very easy to retain the formatting of the original table and I find myself re-applying formatting again and again. Paste Values doesn’t seem to work in the same way it does in Excel. Make this easier, please.
  3. Once again, how about a tables format painter? One grudge I have is that it doesn’t seem possible to replicate the colour and style of a table border. Say I have a table with some nice borders and I add some new columns. I want those borders to have the same borders as the rest of the table. But what was the exact colour and style used? You don’t seem to be able to find out, leaving you to have to try and guess.
  4. Again, the drawing guides, while working well for other objects, don’t seem to operate in the same way with a table. I can’t rely on the slide design and the position of the placeholder. My slides are way to complex and diverse for that. But I need to be able to line things up properly. One of the simplest ways of making a slide look better is to get everything lined up accurately. But PowerPoint doesn’t make it easy.
  5. Finally on both tables and charts. Is there a way we can align them better, so that if there’s a table and chart next to each other that have to be aligned, we are not constantly guessing  to try and match them up?

Anyway, now that I’ve got that out of my system, Happy Christmas to all of you, and happy charting in 2015!

PS – any chance of a macro recorder? Please…?

 

 

 

Visuals should come as standard

Charted visualsWhen I started in market research, the main deliverable we were passing to my very first client was a set of tables, one for each country we were researching for them. The client, a specialist research buyer within an insight team, would then check them thoroughly, trawl through them for mistakes and often find one or two, then use them internally to create whatever reports were needed. Once a year, all the findings would be brought together in a PowerPoint set which would be used as the basis for an insight presentation.

Who sends tables to clients these days? The nature of the research deliverable has changed, and in my view not just what we deliver to clients, but what is used by researchers to report on results. More and more it seems to me that between the production of tables and the stage at which someone sits down and actually starts trying to work out exactly what the research is saying, there is now an intermediary stage which involves the production of large numbers of PowerPoint slides. The analysis and interpretation of quantitative research results is now most often done working from a set of slides than from a set of tables.

Previously, charts and tables were drawn up to illustrate some aspect of the story as identified in the tables. Now, the tables are more often used as a back-up, to drill down into demographics, for example, when an explanation is needed. The actual results are viewed and comprehended by means of scaled visuals, where the difference between figures, the change over time, etc, can been seen in terms of the size of two bars or the shape of a line.

What does this mean in practice? It means that a lot of PowerPoint charting is done which is never actually presented, and a good deal of which might never actually make it clientside at all. It means that there is now an additional, mandatory, phase between final tables and interpretation and reporting during which the dataset is turned into a visual format. It means a lot of toggling between PowerPoint decks to copy and paste slides, and the transferring of data from one chart datasheet to another, in order to bring together the actual story.

To do the job of analysing the results, researchers need a set of visuals straight away. I would argue that we should be moving away from seeing tabulated results in themselves as a research deliverable, even when being delivered to a researcher and not the end user. Tables in themselves are merely an intermediary stage of aggregating data into groups and producing percentages, nets, combinations, etc. Research deliverables should include visuals as standard. At the very least that should include questions charted in an intelligent way, and ideally would include plenty more options than that.

And we need to find a more efficient, fast and accurate way of producing these visuals than creating weighty PowerPoint reports from final data sets. Software now exists which can take respondent-level (or aggregated) data and create charts and tables that can be saved so that final or different data files can be run against the same setup. A basic setup can be delivered to the researcher who can then manipulate it to explore the data – or download it into PowerPoint to browse it. Automation can be used to create a simple set of charts, not the final output but the first thing the researcher will look at when starting to understand and interpret the data. This should be something that’s delivered, quickly and simply, at an early stage post fieldwork.

The technology is already there for us to do this. The days when researchers have to copy and paste a load of Excel tabulated data into charts before they can start to have any insight into what the results are saying, really should be well and truly over.

Five things you can do with PowerPoint macros

Automation with PowerPoint macrosI’ve recently discovered how to do a number of things in PowerPoint using inbuilt VBA (Visual Basic), ie by writing PowerPoint macros, which cannot, as far as I know, be done automatically as part of automation software packages. This means that by commissioning a custom automation solution created in VBA, more can be done in terms of customising the look, feel and use of PowerPoint than automation using a software package. At a time when people are becoming much more aware of the importance of good data visualisation and how much more value can be gained from survey research by presenting it well, it’s important to stretch the functionality of PowerPoint as much as we can.
1.    Hyperlinks. VBA can be used to create hyperlinks on every slide (or on the Slide Master) which means, in Slide Show mode, users can navigate easily around the presentation by clicking. Hyperlinks can convert a linear slide-by-slide deck into an interactive resource which users can explore as they wish.
2.    The Slide Master. Each set of slide master slides is called a Design, and can be manipulated in the same way that a slide can. In my experience, Slide Masters are often neglected in survey reports because it’s easy to paste slides in from other presentations and end up with a mammoth number of Slide Master Designs which are impossible to keep track of. It’s easy to inadvertently mess up a slide by tinkering with the Slide Master, because you don’t know which slides are using that design. VBA can bring clarity here, enabling redundant designs to be deleted, reducing large file sizes and producing a manageable number of Slide Master Designs which can be used in a helpful way to avoid people having to make the same formatting changes to their PowerPoint slides again and again.
3.    Changing the order of objects. It’s easy in VBA to send an object to the back, or bring it to the front. This is important when working with any objects that are overlapped, for example a picture as an object background, or colour-coded shapes to highlight figures in a table. Object order can also interfere with the proper working of hyperlinks. No awkward work-arounds required: with VBA you can control this.
4.    Inserting a picture into a shape as “fill”. Why would you want to do this, instead of simply inserting the picture onto the slide? Because you can use the characteristics of the shape. For example if you want your picture to have effects, like shadow, border, bevel, etc, you set up those effects with the shape and they will be retained. Similarly, the shape and exact position of the shape will be retained. So if your picture is a square and you want to show it as a circle, you can create a circle in the PowerPoint and insert the picture as a fill into it. If you want the same shape to appear elsewhere as a heart shape, you can use the same principle without having to create a heart-shaped picture.
5.    Creating gradient effects. This can be done in VBA for any object or series of a chart, and the chart background. You can specify how many gradient stops and the exact colours, plus also matters of finesse such as the exact angle of the gradient. This enables a high degree of control over the visual output, and means that gradient colours can be varied for different brands, segments, sections of the presentation etc.

Multicoloured dream – using colour in a structured way

Using colour - colour-wheelHaving blogged last week that sometimes it’s best to take out the extraneous colour from a report, I now find myself working on one of the most multi-coloured reports I’ve ever seen – and it’s an absolute joy.

The report is for a client in a very graphically-minded industry sector  and so it really matters that it looks striking and professional. An in-house graphic designer has done some considerable preparation for the reporting. There is a colour palette consisting of ten distinct colours that are used for different sections of the report and different products. Each colour has ten different tones. So there are 100 different and distinct colours to choose from, and yet they all go together as they have been selected by someone who knows about colour theory and what goes with what. It’s a very powerful resource.

There is an attractive background which reflects the industry sector and contains graphics that go with all of the different palette colours, which are overlaid on top. Graduations are specified as well, so there is a great deal of consistency in the look and feel, even though a lot of colours are being used.

There are loads of images, representing brands, products and distribution channels – all in full colour. While the icons themselves are colourful, the charts next to them which contain the data tend to be in the neutral shades included in the palette – grey, with some colour coding to pick out the client brand. Within the individual sections, the background colour has been selected to complement the colours of the product image, as well as going with the overall background imagery. So once again you have colour variation and consistency all at once.

Image sizes are specified to be consistent within sections, but they vary between sections to best fit the space available. Over 40 different icons have been designed to match the look and feel of the report, and are available in neutral tones as well as several shades reflecting where they are found in the report. It’s great to look at but also neat and clear. There’s a lot of information but it doesn’t look cluttered because it’s been laid out carefully with spacing optimised, again by someone who knows how to optimise space!

The only thing it doesn’t have in it so far is real data. That’s on its way, so it will be interesting seeing how effective it is at communicating the research story. I expect it will be pretty effective.

Clearly a lot of work has gone into it and that represents and investment of time and money which can’t happen on every project. But if we are concerned about clients getting more out of the research they commission, investing a little time in project-specific graphic design at the outset is surely something that should happen more often.

Information or decoration? How we use colour in reporting

Colour in reporting 1In brand and advertising trackers, it’s quite common to colour charts according to the actual brand colours. For example, the bars and line for Coca Cola would be red and Pepsi blue throughout the presentation. It’s easy to automate this colouring as well, with automation software like E-tabs Enterprise, and in PowerPoint VBA. You can specify each brand colour according to a precise RGB code, and even have the brand colours vary by country. This is certainly eye-catching, and avoids the smart-Alec in the audience saying “Why is Coca Cola coloured blue?…” during your presentation.

I’ve spent quite a bit of time setting up this colouring automation on various projects. However, after looking at a lot of data visualisations from outside of market research, I am wondering whether this is always the best approach. Old PowerPoint (2003 and older) coloured each series of a chart a completely different colour as a default – hence possible gaffes with Coca Cola ending up blue. In newer PowerPoint, the default is shades of the same colour. This makes more sense in a lot of ways. Looking at large-scale visualisations of thousands of bits of information, every variation tells you something different. But if you have a chart where Coke and Pepsi are already labelled, colouring Coke red and Pepsi blue doesn’t actually tell you anything new. If you can already distinguish between them on the chart, there’s no need for them to be different colours.

Removing colour, or rather neutralising it by making chart series all shades of the same colour, opens up opportunities to do more with the overall design. If the colours match the overall theme of the presentation deck, the effect is more unified and pleasing on the eye.

Colour in reporting 2

A striking dark background works when the chart colours are neutrals. You can make the background textural or include photographs, or have a much stronger graphical theme that the presence of lots of clashing colours would detract from. Icons work best when in the same colour, because the mind looks at the shape rather than the colour. I’ve seen white-on-dark versions of brand logos used in this way to great effect.

Colour in reporting 3

Keeping colours as similar as possible when they don’t have to be different also means that colour can be used to add in more information. Brands could be coloured according to their attributes, for example by price band, product type (business or leisure), region, size, ownership, etc etc. So the colour changes are actually communicating something on top of the brand name. Changes and differences can be made to stand out more. In a multi-coloured chart, a red cross and a green tick are less noticeable than if everything else is the same colour. Significant differences, changes over time, comparisons to a benchmark or the highlighting of the best or worst performer, will all have more impact against a more neutral background.

In the examples above, there is actually very little information being displayed, and this is more obvious if you neutralise the colours. It would be feasible to include much more insight into this visual by using colour to tell us something new.

One frequent criticism of market research reporting is that our visuals are very busy and try to convey a lot of information without it being obvious what we should really be looking at. Yet reporting in other sectors manages to display much more data-heavy visuals in a less confusing way. It is possible to present large amounts of information effectively, if it’s well designed and there’s no clutter. So how about this golden rule: If it tells us something we don’t already know, it’s information. If it doesn’t, it’s decoration and should be banished.

Friday Auto Charting: a simple way to get to charts quickly

Friday Auto ChartingI’m using the blog this week for a bit of shameless marketing. A number of people have said to me that they would love to find a way of getting their data into visual format without having to do a whole load of copying and pasting, when they really want to be analysing what the results are saying.

So, I’ve just launched Friday Auto Charting, a service whereby if you send me a set of tables or a raw data file, I will send you back a set of PowerPoint charts for every question.

The charts are based on your design and template – send me a few examples and I’ll do the rest. There will be two chart types, one for categorical questions (eg Yes/No) and one for scale questions (eg Agree/Disagree). If you need more different chart types I can include that as an optional extra. All the charts are fully editable.

I’ll include up to ten subgroups in the charts (although if you really want ten the charts will look pretty crowded!). You just have to let me know which columns they are in the tables or what the variable or definition is in the raw data. I’ll also show base sizes, and will do useful things like group multiple responses together as one question.

In the raw data file I can also add in nets (eg top 2 box scores or new combinations of age breaks) and add these into the charts as required. I will also supply an Excel file which contains all the aggregated scores, for further use.

That’s all included in the basic price which is £200 for charts from tables, and £300 for charts from a raw data file (SPSS). Optional extras start from £50.*

Examples of options extras: If you need particular brands or responses to be particular colours in all charts, I can do that. Ranking – either throughout all the categorical questions or just some of them as specified. Showing statistical significance – I’d generally do this by including the calculation in the chart datasheet. I can be pretty flexible about how you want the significance displayed. I can provide low base warnings on relevant slides, or hide numbers if the bases are too small. I can position brand logos along the axis throughout. In the raw data file I can change what the percentages are based on, if the definitions are straightforward.

In other words, although a set of charts like this is only a step towards a final output rather than a final report in itself, I can help you get a long way there by programmatically doing a lot of those things that take a long time manually.

It’s quick to set up and run, about a day to turn around. If you can get me interim tables or data, I can do all the setup in advance and could be giving you final charts within a couple of hours of getting the final set. Needless to say it’s very quick to rerun something once it’s been set up. It would be very feasible to set up an Auto Chart for a particular project that had some specific requirements and then run it through periodically for different waves or countries.

Looking at all the different approaches to automation, there are two types, push and pull. Push automation starts from the data and produces a chart (or table) based on what’s in the data file or tables. It’s generally quick and easy, but produces something which is quite simple and would need some working up before it becomes a final report. Pull automation starts with the template, which has been created to reflect what the final report needs to look like. An automation program works through the template and pulls the data in as specified. This is for more complex reports, where the requirement for the final output is known. You can do a lot with pull automation in terms of calculations and custom formatting, so the results should be very close to what the final report needs to look like. This is the best approach where there is a hefty report required which can be specified in advance, and is particularly useful where multiple reports are needed.

Which automation approach is the most appropriate depends on the project. The difference in price is that push automation might cost a few hundred pounds while pull automation would generally cost a few thousand.

Friday Auto Charting is simple push automation, designed to be fast, easy and cost-effective, and get you to seeing the results more quickly and without the pain.

(* Prices are exclusive of and subject to VAT)

Is the DIY approach to survey reporting a good thing?

DIY toolsIn a recent webinar, What’s Hot in Market Research, consultant Ray Poynter described the growing trend for what he called “DIY research”. This includes scripting tools such as Survey Monkey, but also specific aspects of the research process. Reporting must surely be one of them. In the same way that people can now easily design their own package holiday by booking their own flights and hotels and specifying exactly what they want, you would think that people who need to access survey research would be interested in being able to access the data directly themselves and create their own reports.

There are an increasing number of tools available by which survey data can be made accessible online in an interactive way so that those with access can explore it themselves. Is this trend a good thing or a bad thing?

On the plus side, some research buyers and users will find the idea of getting direct access to the data themselves very convenient. They can see what they want to see, and don’t have to plough through a 150-page PowerPoint report to get to what they’re interested in. They don’t need to request additional analysis and they can pull off what they want themselves (if the interface is easy to use). They can navigate and make use of the data when and how they want. Potentially, they can access the data much more quickly than waiting for a formal report. This is particularly useful for results that need to be actioned quickly, such as unhappy customers.

On the minus side, not everyone who commissions market research is particularly interested in the data itself: some only want to know what the results mean for their business or area of responsibility, and don’t have the time or inclination to go exploring data sets for themselves. They would hope that the research agency, particularly a consultancy-led agency, would do all this for them and tell them what they need to know in a way which points towards action. Some users may not really have the technical expertise to interrogate a data set and could make mistakes with it. Often, “real-time reporting” sounds great but, depending on the type of survey, what you often need is for the survey to complete in the field before you start to look at the results, to ensure that quotas are filled and response rates maximised.

At the end of the day, one of the things researchers are good at is looking at results, interpreting them, and communicating what they mean, and the need for this is not going to end. However, for some types of survey, enabling direct access will definitely add value and allow more use to be made of the research, making it more real to people and more convenient.

Someone outside the industry might ask why we aren’t doing this as standard these days. An interface like Google Analytics, available for free, is easy and intuitive, provides recently-updated information, enables overview and detailed analysis and is attractive. Why don’t all our surveys have a facility like this? The answer is that most quantitative surveys are ad hoc and have been specifically designed, so an online tool would also need to be specifically configured each time. This seems to work out expensive, particularly for something that looks good, and basically costs more than clients would see the value of.

Two things about that: firstly, as technology improves the costs will come down. Secondly, if the online reporting tool is seen as an add-on to everything else, it will look expensive, but most of the tools available can be used to produce all offline reporting requirements as well. Upload respondent level data into a tool, configure it for online access, create dashboards and graphics appropriate for the different audiences, and use the same tool to create the presentations which are needed for face-to-face debriefs as well. This is the most efficient way forward.

I sometimes wonder how much our offline reporting processes cost if we add up all the time we put into creating monster PowerPoint reports. Anecdotally it sometimes appears to be the most time-consuming part of the entire research process. Reporting on a large project – just getting the data into graphical format – can easily take up more time than updating questionnaires and setting up fieldwork. If we did work this out, we might find that using a DIY tool instead is not actually as expensive as it sounds. It would need to be instead, though. Not as well.

In other words, if we are to offer DIY reporting, and to be able to afford to do so, we should also be using it ourselves. Imagine a world in which, when the fieldwork is complete, instead of a set of tables, the researcher receives a link to a tool where all the data is sitting, configured with weights and nets etc, ready for her to explore and analyse on behalf of the client. She will then be able to evaluate the results and provide the key insight and interpretation using the tool rather than working from a set of tables. In addition, the client also has access to the same data sets, possibly enhanced with dashboard views that illustrate the story that the researcher has identified.

Sound like a better way of working?

For my review of online reporting tools, visit http://www.traceyhill.co.uk/downloads/

 

 

 

 

Going round in circles: exploring circular visualisations

 

Circular visualisation National Geographic

I’ve seen a couple of excellent circular visualisations recently, like this on length of commutes from the world of print – appearing in the April 2014 edition of National Geographic (I can’t find an online equivalent). It tells a great story – the colour chunks give an idea of relative country size and the length of the bars shows us that although London is a nightmare, it’s not as bad as Brussels (or Antwerp).

What’s the advantage here of making a circle? Well really, it just seems to be that you can fit more on a page that way. Spread out in a line the data would need to be cut up, but threaded round you can see everything at a glance.

I did a little more searching and found this blog post with some exploration of how scale survey data could be dealt with in the same way. Not terribly successful in this instance, but with the right design and some thought about how best to order the data, it could potentially result in a compelling overview of findings, and is certainly something that looks really different.

Also doing the rounds online recently was this excellent circular visualisation of global trade flows between the world’s biggest exporters. Here another great reason for using a circle: it can be used to display multiple connections. Each country’s exports and imports with each other are shown to scale with the colours drawing attention to geographic regions. There’s a huge amount to learn from this: for example, I was surprised to see how little the EU trades with the USA – topical given the current development of a possible EU/US trade deal (TTIP).

This kind of visualisation could be used to show brand repertoire for example, ie responses to a question along the lines of what people buy when their normal brand is not available, to show what brands are similar to each other in the eyes of consumers.

Here’s another possible use: an interactive circular visualisation of issues affecting small fisheries management. Multiple factors are categorised, grouped and coloured around the outside, and the priorities of each stakeholder group represented by the area in the middle (colour coded) on which you can view one at a time or everything at once. I have to admit the specifics are a little lost on me, but I can imagine something like this being used to represent the results of research – including qualitative – with the different ratings of stakeholders, demographic cohorts or segments shown in this way. It’s really engaging.