By MADA Senior Lecturer Robbie Napper
Robbie NAPPER: Public transport user experience: How can we make it better and easier by applying principles of interface design?
NAPPER: Hello, I'm Robbie Napper and this is the Industrial Design Podcast, number 6.
NAPPER: Today I'd like to talk to you about three things. I'm going to talk about removing information, I'd like to give you a refresher on organising information, and, critically, the third thing I'll talk about today is how to do a survey of people to see whether your designs are any good. So, this assignment, "Public Transport User Experience" gives you the opportunity to remove unnecessary information. I strongly suggest you do this - you take this opportunity to make life easier for people. We live, as you know, in a very visually polluted environment. There are signs and sounds and smells everywhere. People are overwhelmed by this information, so I don't think that as designers, we should continue to overwhelm them by having even more information, especially when it's unnecessary. I think we should look for opportunities to do this, because we are in a complex environment and the public transport environment is complex and it does require people to make a lot of decisions as they go along.
NAPPER: So how do we remove information? Well, the first thing to ask yourself is, "Is this information redundant?" If you're looking at a map or a bus stop or another one of these environments that we're dealing with in this project, ask yourself that question. For example, in the City Saver map, the tram map that we have in Melbourne shows Swanston Street as this gigantic rainbow road of trams running through the centre of town. Now, in the context of the overall tram system, that's important, because it shows that a tram starts at one end of town and goes through the city and pops out somewhere else. But in the context of the City Saver area, that information is largely irrelevant. Think about the journey from the NGV on St. Kilda Road, up to the university of Melbourne. There are so many trams that you can jump on to make that journey, it's almost irrelevant what number it is, because all but one of them will take you on that precise journey - they do not deviate. So, instead of having a City Saver map that shows all the tram numbers, we should really just show that there are a number of trams going up Swanston Street, any of which will deliver you to that end of town. That would be removing unnecessary information.
NAPPER: Stations and stops, well, think about a better way of showing where a bus or a tram is actually heading. Is it important to know this information? Well, yes it is. But there is other redundant information shown as well. Do we need a number to identify them? We probably do. It seems like a good system. But do we need to know every single cross street, or is that only relevant to particular passengers? Think about the information that you're bombarded with when you get off a train at, say, Parliament or Melbourne Central, and you want to change to a different train. We need to nest this information much better than we do. The London Tube does this very very nicely. You go through a series of choices and each would bring you closer to your final aim. So, if I get off a train in London, I'm confronted with two options: I can exit the station or I can change trains. Decision one. "O.k., I want to change trains." so I walk in this direction indicated by this nice clear arrow. O.k., then, the next question is "Which line do I want to change to?" The lines are colour-coded - hallelujah - so I know, "Well, I want to change to the Piccadilly Line." or "I want to change to the District Line". So I chose my colour, I chose my line. Then, I walk a little bit further, and then finally, "Which direction do I want to go in? Do I want to go northbound or southbound or east or west?" So the nesting of these choices makes it much much easier, rather than the alternative which is to get off a train, to have an extremely complex series of choices, "Right, do you want to go southbound on the Piccadilly Line? Do you want to go eastbound on the District Line? Do you want to exit the station?" All of these things - it's very very difficult. So, the London Tube nests that information very successfully.
NAPPER: Ticketing transactions, we can always remove some information that is redundant from those. The current myki ticket regime presents us with a lot of information and some of it can be stripped away so we can just see the clear choice that we need to make. Simplify wherever possible. And think about vehicle signage too, for those of you working in that work area. Minutes to arrival - is that necessary? Well, I think that's quite helpful. But the names of lines? I don't think so. The names of lines in Melbourne are particularly difficult to deal with and they only apply to trains anyway. So we need a better way of showing - in the context of a particular locality, say, Caulfield - where a bus or a tram or a train is going. We don't need to show the entire network, we just need to show the selected information.
NAPPER: So that brings me on to topic two for today: using those "Five Hat Racks" - L-A-T-C-H - for organising our information, location, etc. I think location is probably going to be a very important one for this project because it deals with transport and transport is effectively the task of moving one person from one location to another. So location will be very important as a way of organising information in this project. We see a lot of information in random arrangement in the current system and this should not be the case. There has to be an underlying system to the arrangement of that information. I think that location-based information is very important in public transport user interface because it gives us the opportunity to, like I said before, strip away unnecessary info, and given the nature of this project, that will be very very helpful. But of course, don't forget the alphabet, the time, and other ways of arranging information.
NAPPER: So, having spoken briefly about that, I'd like to now tell you about doing a survey. You've already done one type of survey in this project, you've done a site survey of what you're going to do for this work, and you've done that quite successfully. The second survey that you'll use is a survey, in the traditional sense, of people. And there are some very important rules for how we conduct surveys at Monash. So let me set out some ideas and some guidelines for doing the survey and you might want to write some of these down. The brief requires you to develop pictograms. Remember that the brief is pushing you towards the removal of written language and the best way around that is to develop pictograms. So, you need to test them out. Test them out in your workgroups of course, test them out in class, and that'll give you a lot of help in refining them. But it will come to a point where you need to test them with the public - the "public" of "public transport". So, I would expect you to use family and friends as your "public". The sample size and type, well, you need to make sure that you survey a variety of people as well and there are some rules which I'll go through now.
NAPPER: Firstly, let's talk about setting out the survey. I think that this would be best done on A4 paper - let's keep it simple. Let's say the limit to the survey should be a double-sided A4 page. If it's too long, people aren't going to be interested. If it's too short, you're not going to get enough information out of it. So, I think around 20 or so questions would be good. Now, critically, the survey needs to have a preamble at the top. This is very very important because it tells participants that their participation in the survey is voluntary - that's important. It also should tell them that the topic is public transport interface design and it should tell them what your work area is, whether you're doing ticket machines or the City Saver map, for example. This will mentally prime your participant and it will help them give their best response as possible. It also levels the playing field. Now, you need to also say that Monash University is responsible for this survey and that any concerns should be directed to me, so you need to put my name on there, and my e-mail address. The preamble is essential, because the university needs to comply with ethical research standards, and yes, this is research, and furthermore, it's research on humans. So we need to be very careful that people are not forced or coerced in any sort of untoward way into participating. Very simply, you need to ask people whether they are willing to do so, and if they're not, you thank them for their time, and you leave it at that. Now, formatting the A4 sheet is something that I know you're capable of doing, and what I visualise is something like a column of pictograms - numbered, of course - and a space for the participant to write what they think the pictogram represents. I don't think you should have multiple choice, I think you should have open answers. I think 20 questions, give or take, would be a good number of questions, maybe 10 per side, maybe an upper limit would be 30 if you've got that many pictograms. Now, critically, we also need to know about the demographics of your respondents. So, in addition to the questions that you have about your pictograms, "Please write what you think this pictogram means." you need to also ask a few demographic questions - it'll help us analyse the information. I think that age, gender, and whether the person uses public transport more than twice a week, will give us quite a good understanding of who we're dealing with. So, once you've got all that laid out, and you've got your refined pictograms on the survey, it's time to conduct the survey. So approach people you know. Make sure that it's anonymous, and don't force the answers. Don't stand over people while they're doing the survey - just hand it to them, stand aside, and they'll hand it back to you. The sample size is also important here. I think you need a minimum of 20 people. Now that's quite easy because you're working in teams of 2 or 3 here, so you'll easily be able to find 20 people. I don't want 20 students or 20 pensioners or 20 of any particular demographic - I want to have a mix, that's very important. You have to get different types of people to do your survey. If your survey is restricted to 19 year old business and economics students, you will get a particular type of answer because those people have a particular set of skills. So, make sure you mix it up a little bit.
NAPPER: O.k., so you've done your survey. Then what? Well, make a spreadsheet. Write down whether they got the answers right or wrong for each particular pictogram, and if they got it wrong, it's very interesting to know what they thought it was. If they thought that your picture of school, or that your picture of a monthly ticket, was actually a cow or a sunshine, that you've got a big problem, so we need to know what these different answers are. Next, I would expect a certain analysis of the data. It's very interesting on the surface to know that "pictogram 1" was very successful and "pictogram 5" was less successful and it had these strange answers, but the analysis will also enable you to see what types of people got what types of things right or wrong. Cross-reference the answers with the demographics. You might find that older people did better. You might find that younger people did better. You might find that females were more successful at a certain type of pictogram. This will really help you refine what you're dealing with because the next step after doing the survey, of course, is to apply your analysis and to reassess the design of those pictograms.
NAPPER: So, to recap, today we've talked briefly about three things. Number 1: removing unnecessary information. There are vast opportunities to do this and I look forward to seeing what you come up with in this project. Number 2: I've reminded you about the "Five Hat Racks". In particular, location used as a way of organising information - I think this will be very useful. And, I've set out some rules and some ideas for conducting the survey which you'll use to improve your pictograms and hopefully get you towards that holy grail of no written language in your interface design. Good luck.