Interface Improvement Cheat Sheet

    Have you ever faced a situation: “Can you do it even better?”, Having at your disposal only your own head and a certain piece of time? There was a desire to improve the existing interface, but did not know where to start? If so, then the article will be useful to you.

    I will say right away: the knowledge is completely taken from the book “User Interface Design. The art of washing an elephant ”by V. Golovach. For me, the material turned out to be so useful and cool that I wanted to share it. If you have already read the book, you will not find anything new, if not, welcome tackle. The article turned out to be pretty big, but it's worth it.

    So, in order to improve a certain part of the interface, the author advises asking yourself pre-prepared blocks of questions in a certain sequence.

    Block one:


    1. Is it possible to accelerate user interaction with this interface?

    2. Where in this interface are places that can produce human errors?
    Can these fragments be changed?

    3. What is not conducive to learning in this interface? What does the user need to know in
    order to successfully interact with this interface? Is there anything in this list
    that the interface itself does not tell the user?

    These three questions need to be asked yourself in turn. If after the answers it is clear that the interface needs to be changed, the remaining questions need to be asked yourself again after the alteration.

    If it was possible to give a negative answer to all three questions, we begin

    The second block of questions:


    4. Do I know anything about users that makes this interface bad?

    5. Does this interface satisfy all user motives I know?

    6. Is this interface compatible with the environment in which users work?

    If everything is fine on these issues, we proceed to check how user tasks are performed in the interface.

    7. Are there tasks that are inefficiently handled by the interface?

    As a rule, it’s enough to say aloud (or even better write) how the user performs all his tasks in this interface (it’s best to write about yourself and not about the abstract user, for example, “From the Document menu, I open the zeta conversion settings window, enter the value 40 in the Number of people field, then I open ... ").

    This check reveals a lot of inconsistencies or simply missing pieces.

    If this happened, we return to the very first question. If not, ask

    Last question:


    8. Is this interface sexy and can it be made sexier?

    As you can see, there are only eight questions and there is nothing particularly terrible in them. There is only one trick: any product has many functions and, accordingly, integral "pieces" of the interface. Questions must be asked for each “piece”.

    Usage example


    For example, we use the program from the Windows delivery - Calculator (normal, non-engineering, mode).



    Divide this interface into fragments for a separate check.

    They are: the main menu, the display of results and the panel itself with numbers. Accordingly, you need to ask yourself 32 questions: 24 for individual fragments of the interface and another 8 for the program as a whole.

    Menu




    Its only non-standard element is the “Number of digits in a group” switch. If you enable it, long numbers will be divided into parts of three digits.

    We start to ask questions:

    1. Is it possible to speed up user interaction with this menu? - Not.
    2. Where in this menu are places that can produce human errors? Can these fragments be changed? - The name of the item “Number of digits in a group” is difficult to make completely understandable. You can, of course, rename it to "Divide long numbers into groups", but it is very long. Maybe the item should be thrown out of the menu by turning on the default division?
    3. What is not conducive to learning in this menu? - If we throw out the element “Number of digits in a group” - nothing.
    4. Do I know anything about users that makes this menu bad? - Not.
    5. Does this menu satisfy all the user motives I know? - Yes.
    6. Is this menu compatible with the environment in which users work? - Yes.
    7. We list the list of all tasks that the user can solve using the menu. It seems to be nothing problematic.
    8. Is this menu sexy? - No, not sexy. The standard cannot be sexy at all. But here it is not necessary.

    Showing output:




    1. Is it possible to speed up user interaction with the output field? “Obviously, yes, because long numbers are slowly scanned by the gaze.” You need to enable the default splitting of long numbers.
    2. Where in this field are places that can produce human errors? Can these fragments be changed? - If the user needs to read the result of the calculations, and not just copy it to another program, displaying long numbers with a solid can cause errors. It is necessary to enable the mode of splitting long numbers by default. In addition, it is useful to increase the size of the numbers in order to improve their intelligibility. The current interface does not help verify the results of their calculations: the only way to self-test is to repeat the calculations and compare the results, which is unreasonably long. Need some kind of self-test mechanism, for example, you can show intermediate results of calculations.
    3. What in this output field is not conducive to learning? - Seems OK.
    4. Do I know anything about users that makes this field bad? - Not.
    5. Does this output field satisfy all user motives I know? - Yes.
    6. Is this field compatible with the environment in which users work? - On monitors with a large number of dots per inch (for example, on many modern laptops), the numbers can be so small that they will be difficult to read. It is worth increasing.
    7. We list the list of all tasks that the user can solve with the help of the block showing the result. It seems to be nothing problematic.
    8. Is this interface sexy? - No, it’s not sexy, because it’s standard, but it doesn’t cost anything to change: for example, increase the size of the numbers or choose a font with specific numbers. Or do both.

    Panel with numbers:




    1. Is it possible to accelerate user interaction with this panel? - Unlikely.
    2. Where in this panel are places that can produce human errors? Can these fragments be changed? - The intelligibility of the multiply and subtract buttons (pictograms * and -) is not very high, which can produce errors. Increase the size of the icons in the buttons of arithmetic operations.
    3. What is not conducive to learning in this panel? - The names of the buttons MC, MR, MS and M + do not tell the user anything if he does not know their purpose. This is normal for the engineering version of the calculator, but unacceptable for the normal one. It is worth increasing the size of the buttons so that the best names fit into them (or even abandon them, because there is still a clipboard). What is the difference between the C button and the CE button? Maybe this CE can be embedded in the output field?
    4. Do I know anything about users that makes this interface bad? - Users obviously use this interface inconsistently (complex calculations still have to be done in the engineering version of the calculator, but for a frequent calculation, a real calculator with large keys giving tactile feedback is more convenient). The obscene terms on the buttons from the previous paragraph are clearly not suitable for the ever-inexperienced users.
    5. Does this interface satisfy all the user motives I know? - Yes.
    6. Is this panel compatible with the environment in which users work? - Not; at least for new monitors with high resolution and small screen size, it does not fit - the controls are too small (their size was optimized during the 15-inch screens by 800x600 pixels).
    7. We list the list of all tasks that the user can solve using the key bar. It seems to be nothing problematic.
    8. Is this panel sexy? “No, not sexy.” However, it is not clear how this can be fixed.

    When asking yourself these questions, you should always remember that there is a risk of giving the wrong, or, more often, incomplete answer.

    “When analyzing the interface of the calculator, I personally missed such a common task as calculating using brackets. Traditional calculators do not know how to execute it (most likely, the traditional design came about when the memory registers needed for the brackets were an unacceptable luxury). The MC, MR, MS, and M + buttons (and the memory usage indicator) were a clumsy attempt to solve this (in particular) problem. For our calculator, such a calculation is not a problem, so in a good way, the MC and MR buttons just need to be replaced ” - note of the author.

    Conclusion


    So, in the program Calculator is at least:

    1. Show the results of calculations in groups of numbers (317543 => 317 543) by default, removing the corresponding menu item.
    2. Increase the size of the digits in the result box.
    3. Increase the intelligibility of mathematical operations buttons.
    4. Nail the memory operations buttons, but insert the buttons for the brackets and do something with the square root button.
    5. Ideally, at startup, ask the OS for a screen resolution and increase the size of all elements if the resolution is too large.
    6. Implement the display of intermediate costing results.
    7. Make the window always floating on top of other windows (setting). This interface has a problem: if you need to make a series of calculations by copying the results to another window, the calculator window will disappear all the time, overlapping the window into which the results are copied. The user will have to spend time every time returning to the calculator window.

    As you can see, eight magic questions in just a few minutes allow you to create a solid list of desired improvements - which, in fact, is required to start the design.
    I can say that no other method provides such a high efficiency.

    Unfortunately, there is an intellectual trap here. The fact that you now know these questions does nothing and does nothing to help. Their knowledge is not enough - they need to be asked. If you do not ask them yourself, the questions will not help you in any way. For this knowledge to work, it must be actively implemented in its project activities (as, in fact, any other knowledge).

    Also popular now: