I have always been frustrated with my TI-nspire calculator but until taking this engineering psychology class I did not have the proper vocabulary or knowledge to adequately express my frustrations with the design. This ‘rant’ may seem petty or oddly-timed considering it was for a calculator that I had to buy for high school math, but as I said, I have only now gained the proper knowledge to discuss my frustrations.
Ti-Nspire CX CAS Graphing calculator taken from Texas Instrument’s website. This is what the calculator looks like
My first issue with the calculator is the buttons. They have many issues but the first issue is that they are tiny. They are much smaller than the average hand would enjoy to press. It makes it complicated to both press the buttons as well as to read the labels on the buttons themselves. Their size forces the font to be small which further adds to the complications of using them. It is very easy to accidentally press another button entirely because the buttons are so close and so small.
My second issue with the buttons is that they are not enjoyable to press. They are shaped in stark contrast to earlier models of calculators made by the same company.
My friend’s calculator made by the same company with differently shaped buttons
If you look at the shape of the buttons on this calculator, you can see that they are concave inward to fit to the shape of a finger. That shape helps make the buttons discoverable as their purpose is to be pressed inwards. Additionally, the act of pressing the button is much more satisfying which makes the entire experience more pleasurable. Lastly, I’d like to point out that the buttons are much larger and much easier to read because the larger size allows for larger signifiers on the buttons.
The buttons have written out or iconic signifiers to help the user discern what each button accomplishes. These signifiers are often helpful when the symbols are recognizable such is the case with addition (+), subtraction (-), multiplication (x), or division (÷). It is much less helpful when the signifier just doesn’t make sense. On the TI-nspire there are many examples of unclear signifiers but I am going to call attention to the one that looks like “☐” . I may mean something to somebody that knows more about math than I do, but to me and my friends, it was puzzling. The signifiers do not work if they do not communicate a clear point to the user.
Another strange quirk of the button design is that Texas Instruments decided to use split buttons. I am not sure why they decided to use that design but I believe that it fell short in several ways. The first issue I have is that these split buttons are unsatisfying to press. I do not just mean that they could feel better to press, I mean that they are not always clear which side of the button registered if one side did at all. The second issue that I have with the buttons is that there is not always a consistent relationship between the operations that each side does. For example, on some of the split buttons there is multiplication and division, addition and subtraction, open parentheses and close parentheses, but on one there is ^ and x^2…okay…I guess that makes some sense. But on the one above that there is = and a menu for trig functions. Those two operations don’t have an easily identifiable relationship. I am not sure why those two functions were included on the same split button.
Apart from the button layout as a whole, I have very specific problems with the buttons that the company did chose to include. Many of the commonly used functions are either hidden in further menus or require two button pushes to access. The process to access that secondary function of certain buttons is:
- The user must find the button that has the secondary function he or she wants
- Then find the button labeled ctrl (which TI was nice enough to color differently)
- Press ctrl
- Then press the original button to use the secondary function.
This design is inefficient. I understand why some functions require two presses. I understand what they were trying to do. My justification on Texas Instrument’s behalf is that the addition of more buttons would have meant that either the calculator would have needed to be bigger, or there would have been more buttons, or more things hidden in separate menus. Therefore, they decided to use secondary features for certain buttons. I understand that choice but in this case I think that they did not put enough thought into it. They chose to hide common functions behind other buttons thus making them inefficient to use despite how commonly they are used. For example, the square root function, ln, pi, e, and theta are all hidden as the secondary use of another button. That decision makes it tedious to use those buttons. There is a relationship between x^2 and the square root of x and that is why the square root is a secondary function of x^2 but because they are equally as important, it becomes tedious to have to press extra buttons to access that functionality. The decision has a reason but the costs vs. balances of that decision were not thought through.
My largest problem with the calculator, however, is the keyboard at the bottom. Taking up approximately one third of the button real-estate on the device, the manufacturer must have thought that users would be pressing the letters at least as frequently as the numbers or functions. From my personal experience both in high school and in college math, that scenario is not even close to the reality. The letters that I type are x,y,z,t,s, and g. I have not used all 26 letters in the same problem and I cannot figure out why TI thinks they all need to be accessible. I would have gladly given up the ability to type full essays on my calculator in exchange for larger buttons or a larger screen. However, I would like to point out that in the choice to include a full keyboard, TI did not include a regular QWERTY keyboard as is the standard for typing. No, they decided to go with an alphabetical keyboard instead with the space bar following the Z. There is nothing familiar to the user about the keyboard and it causes a lot of problems. Again, I have a possible explanation for why they chose to do that. My guess is that they realized that the most popular letters in mathematics are x,y, and z and the only way to make them easily accessible is to use an alphabetical keyboard because X, Y, and Z are at the end of the alphabet in sequential order. I think this claim is true because they even highlighted those three letters to make them stand out. However, this choice was short-sighted. By implementing such an unfamiliar layout, it makes it hard to find those letters anyway and slows the entire process down. They could have used a regular keyboard, highlighted those letters anyway, and also included them larger on other buttons closer to the actual numbers. Previous calculators have done without the dedicated keyboard so I am not sure why this one needed a dedicated keyboard.
However, according to this article, if Texas Instruments did decide that it was necessary to use a keyboard, neither the QWERTY nor the Dvorak keyboard would be the best choice, let alone the alphabetical keyboard that they did use. According to this article, a keyboard layout like this:
Keyboard II from article 1
is much faster and more efficient (fewer errors) than the other keyboards for typing when used with one hand as is often the use for both phones and keyboards. Therefore, if TI intended to include a functional keyboard on this calculator, one with this layout makes more sense for their intentions.
However, in a different study, it was found that an entirely different keyboard layout would be more suitable for mobile devices such as a PDA, phone, or calculator than the QWERTY or Dvorak keyboard and possibly even the one mentioned above. Here is the keyboard that was the best rated in measures like accuracy, time, and likability:
Keyboard that is best for mobile devices like a calculator.
It may be apparent that I wrote this post after doing seven hours of calculus homework and becoming increasingly frustrated with my calculator. I think that I am pretty harsh on TI for designing such an ergonomically inconsistent and frustrating calculator. However, I think that they make good tools they just don’t make them for their users. If they could design their calculators for the task or even for the human (or ideally both) then their users would be happier spending so much money on their products. TI only needs to make a few small changes to make a world of difference. It just adds up.
Li, Chen, & Goonetilleke. (2006). A heuristic-based approach to optimize keyboard design for single-finger keying applications. International Journal of Industrial Ergonomics, 36(8), 695-704.
Hsiao, Wu, & Chen. (2014). Design and evaluation of small, linear QWERTY keyboards. Applied Ergonomics, 45(3), 655-662.