Thread Closed 
WEEK 12: How Different Groups Spend Their Day
11-13-2014, 11:46 AM
Post: #1
WEEK 12: How Different Groups Spend Their Day
WEEK 12: How Different Groups Spend Their Day

Here is the link of the visualization:
http://www.nytimes.com/interactive/2009/....html?_r=0

This visualization is about how different groups spend their day, using a survey which asks thousands of American residents to recall every minute of a day by The American Time. As you can see in this visualization, the x-coordinate means every minute in a day, and the y-coordinate means the percentage, And combine with x-coordinate and y-coordinate, we can know what percentage for a specific activity the American are doing on a specific time.

And here are some point i like this visualization:
1, I do think from this visualization, i can clear see the general layout for in a specific time, what the main activity the American are doing, for instance, from the 'Everyone' option, i can clear see, most American are working from 6 a.m. to 6 p.m., and most American are sleep from midnight to 6 a.m. and so on.
2, When you point at some point on this visualization using your mouse, it will show a tooltip with detailed information, And if you click a specific point which you want to know, it will shows more detailed information, for example, if i want to know what percentage the American are working at 3 p.m., so when i point at the '3 p.m.' axis in the working region, it will shows me 'At 3 p.m., 29% of all Americans are working', and the i click once, it will shows me the average time spent per day for everyone, men and women.
3, There are many options at the left top of the visualization, which including almost all categories of all American, so, i can select what i interested in.
4, The colors this visualization used are easy to distinguished, which avoid mixed colors.
5, In this visualization, it only marked main activity with large region, for the activity with small region, although it did not marked, but i can use mouse to get detailed information for what i need,, for this, it avoid the overlap parts, to make readers clearly see the result.
Find all posts by this user
11-13-2014, 08:48 PM
Post: #2
RE: WEEK 12: How Different Groups Spend Their Day
This graphic is a stacked graph. I think the purpose is wholly exploratory. I think it does a good job allowing the user to filter the information across many different attributes. I had fun playing around with the data. The interaction is what makes this particular piece so effective. On mouseover the region tells you the percentage of americans in that filter that are performing that activity at that time. This is cool, because it is hard to tell absolute percentage from just looking at the graphic, but with the mouseover it becomes extremely clear. I think the visualization could be improved by allowing the user to filter across multiple columns. For example I wanted to know what males aged 15-24 were doing with their days but I could only filter on one of these criteria. Since the purpose of the graphic is exploration I think that the difficulty of seeing absolute changes between the categories is less important, which is the major drawback of the stacked graph. All in all I really like this Visualization.
Find all posts by this user
11-16-2014, 12:29 AM
Post: #3
RE: WEEK 12: How Different Groups Spend Their Day
Stacked graphs proves to be an excellent medium of representing data which changes over time. Some of the points which I think could have aided in improving the overall visualization design

1. As correctly pointed out by nbertagnolli adding a functionality to filter the data across to view data across multiple attributes is a great suggestion. Another feature which they could have added is filter the content across different time period. For Eg, If we wanted to view percentage of activities carried out by a group within a particular time range say, (9 pm to 11 pm).

2. Tool tip is indeed a great way to view the additional data information. However adding an information box in order to view statistical data based on selected time would provide an excellent summary indicating different percentages of activities carried out on a particular time period.
Find all posts by this user
11-16-2014, 07:51 PM
Post: #4
RE: WEEK 12: How Different Groups Spend Their Day
This visualization is very neat.

It shows everything that I would like to know in this context. I love the categories of different individuals (age-group, ethnicity etc.) on the top left corner.

I very much like the feature to single out one particular activity by clicking on it and see it's distribution through out the entire day.

One layer that, I would like to see on top of it is the distribution of activities of individuals from different cities across the globe through out the day.
Find all posts by this user
11-16-2014, 10:21 PM
Post: #5
RE: WEEK 12: How Different Groups Spend Their Day
First off, I have to say I love the amount of interactivity in this stacked chart. That being said, I have to say that charts stacked in the manner in which this one is are not very good for determining actual data values. For example, if you were to quickly glance at the data for noon for the work category of this chart, you'd tend to think that at noon ~70% of Americans are working. This of course is not true, as if you rollover the actual data, you can see that only 28% of Americans are working at noon. This initial determination that at noon 70% of Americans are at work is fueled by the y-axis's labeling, and without further analysis one may not notice that the actual baseline for the curves are not uniform.
The non-uniformity of the baseline leads to further problems even if one knows that the baselines are not uniform, as the distortion of the baseline makes it hard to accurately determine data values.

Luckily, the amount of interactivity in this chart, largely negates all of the negatives mention above. The fact that upon rollover of a category the actual values for the category are displayed, and the fact that you can select each category displaying a bar chart of it's data(with a uniform baseline), completely help minimize all of the issues one could ever have with a stacked chart and as such make this one of the better stacked charts I have ever seen.
Find all posts by this user
11-16-2014, 10:54 PM
Post: #6
RE: WEEK 12: How Different Groups Spend Their Day
This visualization really captured my attention and was fun to play around with. The amount of interactivity in this chart makes it interesting.
The stack chart efficiently uses the category feature to minimize data and focus on the data being viewed presently.Also the feature to single out any particular activity is nice.
Overall a very nice visualization which effectively delivers the message it want to convey.
Find all posts by this user
11-16-2014, 11:06 PM
Post: #7
RE: WEEK 12: How Different Groups Spend Their Day
A very detailed and interesting piece of visualization.

The interactivity gives a lot of the required information. The percentage data is highlighted when we roll over which is very informative. And also upon clicking a particular part of the stack graph, we get a detailed information about the time spent by that set of people doing that particular work like eating,sleeping etc., whichever activity that we would've selected. I liked how detailed this information is. Not only is there a small explaination, it also shows how many minutes is being spent on that activity. And once a particular activity is chosen, we can select any of the group of people to see how much time they spend doing that activity in a day. This helps in better comparision. This could've been taken a step further by allowing the users to select their choice of group of people at a time and compare the time spent on a particular activity.
I found this visualization very useful and informative. It utilizes the given space well and also I feel the color palette is used very well.
Find all posts by this user
11-16-2014, 11:24 PM
Post: #8
RE: WEEK 12: How Different Groups Spend Their Day
I think this visualization does a good job of packing a lot of data into a flashy visualization. It does a good job of managing the plethora of data without cramming it in all at once through the use of a good filter. The various filter options provide the user with a number of views so they can gain insight on the data. In addition the slider helps to point out various facts. Initially it can be a little confusing to determine how the geometry is working and how it relates to the axes but after some brief analysis it makes sense.
Find all posts by this user
11-17-2014, 12:29 AM
Post: #9
RE: WEEK 12: How Different Groups Spend Their Day
This is a very interesting and detailed piece of visualization which gives a sneak peek into how different groups of American people spend different times of their day doing different kinds of work.

The main view categorizes how everyone spends their time. Clicking on a specific group in the top right corner shows how a specific group like 'Men' spend their days.
Also their clicking on any activity isolates out the statistics of the main activity. In order to get a more detailed view one can even see how a particular group spends what time on a particular activity.
The interactivity of the graph is nicely done and one can hover over the graph to see the exact percentage of people spending their time on a particular work.

The only downside is that in the main view the stacked area chart doesn't really help a user understand about the percentage of people doing a specific activity.
Find all posts by this user
11-17-2014, 10:50 AM
Post: #10
RE: WEEK 12: How Different Groups Spend Their Day
The critique is excellent.

This visualization is essentially a stacked time plot for multiple time series. Each time series represent the percentage of people among a specific group performing certain activity. For example, the orange band represents the percentage of people who are working.

Since the time series are stacked, the value of the time series at each time point is represented by a vertical span. This encoding is unaligned spatial position. This strategy works well for time series with large spans, like "work"". However, it performs poorly for time series with small spans, like "job search" (the 6th band from the top). The author use interactivity to resolve this problem. However, since the band is so thin, it is still difficult to locate the band. Maybe we could put some legend on the side and by clicking the legend we can select the corresponding band.
Find all posts by this user
Thread Closed 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)