This might be difficult as the amount of counterbalancing applied might not exactly counterbalance the ordering effects.

Internal and external validity sound like they are extremely difficult to achieve at the same time - less control will decrease internal validity, and more control will increase external validity.

This is probably why good experiments need to have a lot of data to give as much variation to increase external validity, while reducing the possible causes to the independent variables we introduced.

Wow, this is really interesting. Do people really recognize the reverse-text direction as backwards?

Unfortunate incident of miscommunication and having no one speaking Welsh in the team, or even taking a step further to at least run the text through a translator!

The tone and implied meaning in words in different languages are extremely difficult to be familiar with.

Many of the gamers would have seen this when we repeatedly press shift in games, and a pop-up window comes out to ask if we want to enable sticky keys.

Difference between subtitles and closed captioning is that subtitles only include dialogues, while closed captioning also includes other sounds such as engine starting or plates dropping.

One surprisingly common disorder is dyslexia, which affects nearly one in five people.

Here is an informative video: https://youtu.be/zafiGBrFkRM

Monospace fonts work well with codes, as codes are very structured and well-aligned codes are much easier to read. It also helps when dealing with strings, as comparing strings become a simple task, as position of individual character does not differ according to the content.

I think short lines help with saccades, as it often becomes difficult to track the start of the next line if one line is long. In this reading, it is difficult to go to the next line for the longer lines.

Would percentage < 100% make two lines overlap? Or does CSS automatically adjust the setting to make sure that the two lines of text stays separate?

Sometimes I feel that dark backgrounds are nicer, as they are not too bright and less tiring to look at for longer period of time. Bright backgrounds, in my opinion, make it harder to focus for a long time.

Top 2 examples (I think)

Saturated color receptors have difficulty seeing the complementary color

We don't separate a color into RGB components, or CMYK components. We usually see a color family, and distinguish which shade a color is.

Is the spam flag the leftmost column?

Are textual differences considered differences in shape too?

Table below uses position variable to display data. It is easy to focus on one row or column of data, thus having selectivity, and it is also easy to ignore irrelevant row or column, thus showing associativity.

Wouldn't this be considered inefficient? From the look of the remote, there seems to be no channel change or volume change buttons, which are the most frequently used buttons on a remote. Having to have more steps to achieve the goal seems like a bad design to me. I think the reduction went too far.

It happens when there is no space to move the mouse, or the finger in the touchpad case

In Windows, there is a setting to avoid the repainting a window, and rather just show the bounding box of the window as it is dragged. I think it is a similar concept.

I guess this means that the raw events will not be delivered, only the higher-level translated events. Is that right?

Why is this so? Is it because it is difficult to set a good default value for objects?

I think it might be reasonable to set a visible size, if it has to be changed anyway, to avoid the mistake mentioned here.

I've heard a similar concept called "depth", as in PowerPoint where we can send objects behind or in front of others.

Why is this so? Wouldn't it be better to note the problems the user face, and then guide them later in order to observe further problems?

What is the main difference between heuristic evaluation and user testing in terms of involvement of the designers?

It can be good as a common problem found by many evaluators indicates that it is a rather serious problem.

This might be why user testing is so important. Designers think very carefully about the interface, and know a lot about it, and that separates them from the users by far.

Bad color scheme of the table title (Visibility/Graphic Design)

Specificity helps a lot, as even the most obvious problems might slip our minds if we attempt to fix it later.

I think this means that the evaluation should focus on what is already there, not what can be added.

I think the example on the left is different from what is mentioned in the text.

What's the advantage of this system? I can't really think of a situation where a child outside a parent is useful.

The two links in the Google Form are inaccessible :(

I actually check the status bar before clicking on a link that I don't know where it will lead me to. I guess different users show different loci of attention!

Adding "estimated time" will be helpful in this case. It is difficult to approximate the time left from a progress bar.

https://www.youtube.com/watch?v=FV97j-z3B7U

This video shows the different fps at 5, 10, 15 and 30. To me, it the statistics sound about right - for 10 fps and above, the ball looks to be in motion, but for 5 fps, the ball jumps between two places.

It is difficult to decide when to step in, as if we step in too early, we might miss out on precious data, but if we are too late, the user might feel uncomfortable and uneasy.

This helped a lot during our paper prototype testing. It is really difficult to know what the users think without their explicit input, and we have to resort to guessing at the end.

I think in order to adhere to these rules, a lot of preparation should be done by the designers. A unprepared test not only does not produce much meaningful results, but also might make the user uncomfortable!

Also, another way to look at this is, in my opinion,: "there is always a room for improvement"

Is this feedback from the users? If so, I think it kind of contradicts with statements in the following paragraphs where it says users are only involved in two places.

I think this is because users find it difficult to elaborate what exactly is wrong or could be better, while it is easier to compare and contrast. When the designers come up with alternatives first, users can easily say which one is better, or what could be improved.

This is a little bit confusing. I think it will be better if it is changed to "top" and "bottom"!

In low-fi version, the only change that can be made is the radio button, which will not cause any change at all. For low-fi version, we can test and get feedback on the "look" of the prototype.

In the high-fi version, there are a lot more things that can be done, and we can get feedback on the actual usage experience of the prototype.

I think building personas will be great for initial ideation stage, but as the design process goes on, it might limit our view to satisfy only a small group of users if not used carefully.

It will be important to keep the suggestions short, as users will probably not read a long message

First, user probably didn't know he/she was typing into the font size box. Second, the font size change will be immediately visible if successful - there is no need for an error message, because if no change occurs, user will take appropriate action.

A small typo - missing space

Wouldn't this be a bad mode design? If we can eliminate repetitions in two different modes, we can surely combine them into a single mode, eliminating the need for mode switching.

This sounds quite similar yet slightly different to the definition of description slip. 'Starting in the same way' could be seen as similarity in the action. Could capture errors be seen as a part of description errors? Or how are they different?

Is this the reading for Learnability 1? I can't seem to find any mention of Gimp there.

The mouse has to move far in a very small margin of error

This is especially true in some web pages, If we move the pointer to the submenu before the submenu's appearing animation finishes, the main menu closes thinking that we moved our pointer away from the menu.

Why is aggregation part of the answer? It does not seem to me that users deal with multiple URLs at once, and most browsers do not allow selection of multiple URLs.

An example that utilizes temporal locality

It can help with learnability. Letting the users know what other users have typed allows learning by watching.

Also, it can reduce typos, increasing the safety of the system.

I think this is a good practice, not only because it is easy to change the default text, but also it draws the user's attention to check if the contents are right, and guide them to type in a new, more proper name

How can designers then efficiently propose a new, good design that is affordable? For example, how did scrollbars gain their "meaning" that almost all users nowadays are familiar with?

The left-turn signal is turned on by moving the stalk down, and right-turn signal is turned on by moving the stalk up - up and down matches how the hand moves while turning the handle towards the intended direction.