Category Archives: HCI

Controlling Number Entry Using Sifteo Cubes

As you have probably gathered, I am really interested in Number Entry. I’ve been thinking a lot about how we enter numbers – whether we think of them as a series of digits, or as full numbers. This has implications for the way we might ask people to enter numbers on an interface.

I was given an opportunity to test this out recently at the CHI+MED interface hack day. Last Friday (9/8/13) a group from CHI+MED came together at Swansea university to hack some interfaces to investigate novel number entry methods.

I got a chance to play with some Sifteo Cubes. Sifteo Cubes are fun little blocks with screens that are aware of which other blocks they are next to. This allows for some fun physical interaction, with the user picking up and moving the blocks around.

At a previous idea generation session, we came up with the idea that you might want to let a user enter numbers on the sifteo blocks using both a digit and number strategy. The difference between the two meaning that you could either control the number digit-by-digit (that is, incrementing the digit 9 in the number 659 would result in 650) or by controlling the whole number (that is incrementing the digit 9 in 659 would result in 660).

Using one Sifteo block as the “controller” and the other two as the numbers, I created a system to allow the user to enter numbers using any strategy they like. See the video for an example. You can see that they can control each block separately. When the blocks are joined, the entire number can be incremented (by placing the control block to the left or right) or just the digit can be changed (by placing the control block above or below).

Programming the Sifteo blocks was an interesting challenge. You need to program in C++ (a language I haven’t looked at for a few years) and at first, reading the example code was a bit daunting. But after a while I managed to hack it together. The really strange thing about doing this was working with images. Unlike many programs where you can directly write text (or numbers) to the screen, when programming for the Sifteo cubes you are dealing with lots of static images, and swapping them in and out as you need. Meaning in my application in the video, I have 10 different image files, one for each digit.

I enjoyed the experience of hacking and playing with the blocks. I think there are some interesting questions that could be explored using them. We will see!

Tagged , , ,

Ocean’s Eleven-ing the Bank

In a move that will only cause me to get abuse from my colleagues, I’m about to tell you about a super fun little maths problem I’ve been working on this evening. To make it even more likely that I will be called a nerd/geek, this problem was inspired by a photo of a number pad I took a few days ago. Here is the photo, and here is the story.

This photo was taken in a bank. It was also taken with a lot of espionage skill. It’s the number pad on the back of the cashpoint. This is inside the bank, and will be used when the cash in the machine needs to be stocked. If you wanted to steal the money from the machine by going in this way, you’d need to know the PIN. Which makes the security of that PIN pretty important.

Now, I know the photo is blurry (I didn’t have time to take a decent photo else they’d have thought I was casing the joint) but you can see that some of the keys are brighter than others. This is an example of a Desire Path (which is a post I still need to write about). A desire path shows you how people really truly want to use an object. The cleaner keys show you that these are the ones getting regular use, the dirty keys are not getting pressed as much. The reason this is interesting is that I now know which numbers are used in the PIN of this entry system. Awesome.

Now this made me started thinking about the security of such a system. If you, for whatever reason, can’t avoid giving away which numbers you’ve pressed on one of these key pads, how can you make sure you’re choosing the most secure set of numbers? Just choose any four numbers and hope that the fact the person trying to crack it can’t tell which order you typed them in? Well, as it turns out, no that’s not the best strategy.

If you know that my PIN involves the numbers 1, 2, 3 and 4 then that tells you my PIN is one of 24 possible different combinations, because the order you use those numbers matters. My PIN might be 1234 or 4321 or 1243 or 2314 and so on. So choosing 4 different numbers for the PIN for this machine in the picture means a potential thief would have to guess up to 24 different combinations. Which actually, isn’t that secure.

This is where the maths comes in, what happens when the PIN involves a repeated number? In this case, only 3 of the keys on the number pad would be clean, rather than 4, as one key would be pressed twice. For instance, if the PIN was 1233, only the number 1, 2 and 3 would be clean. It may seem counter intuitive, but this approach actually increases the security of the PIN in this situation, even though it uses fewer numbers. The reason is it adds an element of uncertainty to the mix, there are now three different sets of four numbers that coud be used in the PIN. If 1, 2 and 3 are clean, the numbers in the PIN could be 1233, or 1223 or 1123. Even though there are fewer ways to arrange these numbers, because there are three is actually means that knowing the three numbers used in a four digit PIN means the thief would have to try up to 36 different combinations of numbers.

So my advice to the bank, after doing this maths is that they should pick a PIN with only three different numbers in it, not one with four. Or they should just clean the number pad, it looks kind of gross.

(The maths in the post was a series of permutation calculations, I’ve saved you from having to read them but if you’re interested I can share. Maths is cool.)



Tagged ,

Other people who think that number frequncy research is cool

Today’s xkcd comic made me oh so happy. Mainly because Randal Munroe (the comic’s creator) and I are clearly incredibly cool people who are fun to talk to at parties.


As someone who never sets their alarm to a number ending in 0 or 5 (I woke up at 06:56 this morning) I can sympathise with the character in the comic. But more importantly, this is an example of number frequency information being used for humour! Yes!

It reminded me of the one other time I have seen this done. In the TV series Nathan Barley the title character extolls the virtues of his Wasp T12 Speechtool phone. Aside from built in decks and anti-shark technology, “It’s got a massive number 5 because thats the most common number.”*


Now I’m not saying that this phone is the entire reason for my most recent research. But. Well. Maybe it is*?

*Almost definitely not true but hard to prove. Also look at this asterisk efficiency.

Tagged ,

Gender Discrimination at CHI 2012

CHI this year was incredible. Amazing. It was my first international conference and wow, what an experience it was. I saw so many fantastic and thought provoking presentations, from the brilliant ones on text entry (my one true academic love) to a fascinating presentation on making interactions purposefully uncomfortable. Everything was very thought provoking. And the parties afterwards. Well. Austin is the most partyful town around. The evenings were unforgettable whilst at the same time being slightly hazy the next day… Overall, it was, as one house in Austin put it: lovely.

Now, with that stellar review in mind, I want to talk about the less awesome side of CHI this year, and that was the feeling of gender discrimination running through the event. This was something that I began to notice myself, and then began picking up on when reading the CHI related tweets (the twitter community made this event even better, reporting on interesting talks when you were unsure about where to go, and letting the conference venue know that the air conditioning needed turning down!) Let me tell you the story through a series of tweets. [All the tweets used here are from public accounts and, when not from me, have had user details removed.]

Before attending CHI, my supervisor had asked me if I was going to the Women’s Lunch. I said I wasn’t. I am keen on supporting women in fields where they are under represented, I was a member of the Hoppers society in Edinburgh and helped organise events to offer support to female computer science students and also invited female speakers to talk about a career in industry. I felt these events were necessary and justifiable because females were so under represented on our course. And we regularly heard tell of discrimination against the women doing computer science. The drop out rate for females on the course was significantly higher than for males. So in these situations I believe that providing support and gender specific events is important. However, in domains where there is not a clear divide, I think that holding events specifically for women is hard to justify, and discounts from fields where this support is necessary. That’s how I felt about the HCI community, there are plenty of women in the office here at UCL, and were plenty in the office at York. Therefore, in my eyes there was no problem, and so no, I would not be attending the Women’s Lunch at CHI.

Now, at the conference. I had a paper accepted into the Text Entry workshop. Awesome! I was very excited about this. When I turned up, I was the only female in a room of around 25 people. This was unexpected. I was not made to feel different or excluded, and felt part of the group, but I didn’t think I’d be the only female there. Meeting up with my supervisor later that day I admitted that maybe the number of women in HCI perhaps wasn’t quite the same as the number of men.

At the opening keynote I was to find out! We were presented with a breakdown of the people who were attending CHI. One pie chart showed the number of students, academics, and industry types that were attending this year. Very interesting. The next chart showed the gender split. From memory, the split was around 60% male, 39% female and 1% not answered.

This was fine, interesting graphs, good to know who’s attending, made me realise the small number of females. Now, what happened next was not so fine. The Conference Chair, Joseph Konstan was up next. He took the stage wearing an oversized cowboy hat, which covered  his eyes in shade, so we could only see his mouth. He decided to make a joke about the pie chart, he said something along the lines of “maybe next year we can run a session to help those people work out what they are!”. I was shocked. Now I don’t know who was in that 1% of people who didn’t answer, it could be people choosing not to divulge that information, or it could be people who feel that of the two options: male or female, neither suited them. But to come on stage and say he would “help them find themselves”, that showed a level of incredible ignorance, and could definitely have been very insulting to some people. I wasn’t the only one to notice.

This was when I became aware of people reporting these sorts of things on twitter. And the fact that I had to start collecting the tweets. Like sexist pokemon.

One person pointed out that many of the examples being used in talks, when needing to talk about a novice user were female, the “soccer moms” and the old ladies. No male novice users out there apparently.

Some others pointed at the way that women were being treated at the conference, one person letting men know that if a woman shows any interest in your work, it’s best not to tell that women that you think she may not be smart enough to grasp the content.

Another person pointed out how, in one session, a female first author was introduced as “talking on behalf of her coauthors” whereas all the male speakers in that session were simply introduced, even if they weren’t the first author.

Then we take a look at the videos presented this year. A couple were quite interesting. One, for a device called TEROOS, showed a robot on a woman’s shoulder, being controlled by a man. Whereas another showed a women stressing out over having to cook a meal, everything was going wrong but with a smart cooking system, she was able to make the meal successfully in the end. That was fine. But then we find out she has been stressing out over a meal she is cooking for a man. She places it on the table in front of him then watches him eat. She has not made anything for herself, just the man. In fact at the end she has to ask the man if she is allowed to have a bite of the meal she’s made. Thankfully she is allowed some, and doesn’t starve.

All this. In the space of one week. Maybe I was wrong about the discrimination at CHI. It’s there alright. And I know, sometimes it’s symbolic (in the case of the videos) but there were still reports of women being made to feel inferior due to their gender, whilst at the conference itself. And this just isn’t good enough.

By the end of the conference, I think that this had come to the attention of the organisers, and when Joseph Konstan got up to give the closing speech, he apologised. But he didn’t really apologise. The message he gave was something along the lines of “I’m sorry if you were offended”. As in, ‘I’m sorry if you chose to be offended by this thing I said that I do not believe to be offensive’. Meh. He talked about “slips of the tongue” which I think can’t justify the prevalence of these sorts of comments. These aren’t just slips of the tongue, they represent an ingrained and malformed beliefs that need to be addressed if we are to feel like this community is an equal one.

The one shining light at the end of the tunnel came from Hugh Herr. Hugh Herr gave the closing keynote. And what an inspiring keynote it was, about the fantastic work his lab at MIT is doing for amputees. During his presentation, he spoke of how humans were designed, and then spoke about what the smartest human in the world would do, how she would design a human. Fantastic. There you go. Hugh Herr’s hypothetical smartest human is a woman. There, that wasn’t hard.

I don’t want this post to seem like I was bitter at CHI, or that I didn’t feel welcomed into the community. Because I did feel welcome, I really did. But the fact is that mine wasn’t a universal experience. It is clear that in some areas, some men are still making everyone else feel like second class citizens there, and sometimes joking at their expense. And this isn’t good enough. I’m not suggesting that every talk at CHI next year lists women as the superior gender, all I’m asking for is some equality. Something needs to change.

Tagged ,

Audible thinks you want to keep your books a secret

Interestingly, in relation to this post I made a while back about advertising the book you’re reading on you kindle, audible’s new spotify adverts suggest you might want to download an audio book so that you can listen to book you might not be so proud about reading; thus meaning no one else around you needs to know that you’re enjoying the works of Stephanie Meyer.

It seems the kindle does not deliver for me in terms of not sharing enough information, nor does it deliver for audible listeners in terms of it giving away too much information! Poor kindle.

Tagged , , ,

Comedy on the radio

Here’s a link to the programme Word Of Mouth. This programme was about communicating research using comedy. This theme meant that they came along to Bright Club on the night I was performing and asked me a few questions before and after my set. They interview me to find out what I thought of the experience, and how I found it overall. I appear at the beginning and the end of the programme. At the end, they interviewed me just as I came off the stage, so I sound quite high on adrenaline!

Have a listen to the programme, the bits where I’m not talking are also pretty good too ha.

Tagged ,