Feedback on other blogs

1. Group 10: Ubuntu
   Home | UBUNTU DATA VISUALS (wixsite.com)

First of all, the blog looks very interesting, clean and professional! In the following feedback, both positive (‘great designs’) and doubtful (‘remarks’) aspects are discussed.

Great designs:
- The emerge phase is very elaborate with many different conceptual ideas. Of these, the honeycomb state map is a clever idea that is self-comprehensible and beautiful to look at. It shows a good interplay of different variables using colour and line thickness.
- The flower map is something I fell in love with! I would really recommend using it in the final design. Don’t let anyone change your mind about that! I do have a small remark regarding this visualization (see remarks).
- The hierarchical scatter plot is an easy but clear idea. It basically represents different panels that each show a subset of the data.

Remarks:
- The line plot that combines the number of police killings with the variables share pop. black, median personal income, poverty rate, unempl rate, and education looks interesting. However, it may now appear that the different levels for the variables also indicate a larger relative importance of that variable compared to the others. This is because no scales are provided on the axes. Therefore, a higher level for one variable might now result from a larger unit scale associated with that variable. I think it would be benificial to provide the unit scale next to the axes, or to standardize the variables if possible.
- The dot plot might look a bit simple on its own. Maybe it is worth rethink another idea for that when it would not be combined with another visualization.
- I am not sure how many killings were recorded for each state, but my prejudice alarms me that this number might reach a substantial magnitude. Therefore I am a little concerned whether the petal representation for each killing might not ‘over kill’ the clarity of the visualization. Maybe it would be good if one petal represents 5, or 10, instead of one single petal. Of course there could be some issues with rounding. Maybe in that case you could show only a part of the petal (if there would be 7 killings, this would result in 1 petal + 2/5 petal).

— — — — — — — — — — — — — — — — — —— — — — — — — —

2. Group 16: Group 1
Pattern of Intergenerational Supports

At a first glance, the blog seems visually appealing and the content published up to now is quite informative. Background information is well-stated and research questions are clearly described. The data is explained in detail and initial exploration is satisfactory. Last but not least, design phase seems to be conducted in an efficient way. Below are some recommendations for 2 things each to keep and to change.

2 things to keep:

1. The visualization similar to veins (B) is definitely a creative way to represent the relationship between parents and children. Different colors in directions and different size in levels are the concepts that are easy to understand. It is also utilized in the emerge phase. So, it would be good to keep this representation and work further on it.
2. Interactive dashboard to show provincial heterogeneity, which is created on the emerge phase (1), could be really useful and easy to understand representation of the overall situation. So, in general, any interactive visualization with the country map should be very informative for the big picture and should be kept.

2 things to change:

1. The pie chart that represents different age groups by sex (A) seems a nice way for comparison. However, dividing sex in a single pie of age group might make the interpretation complicated. It might be difficult to compare for instance women in different age groups, because the information for sex is not adjacent. Emerge phase for it shows a promising improvement. First layer could be a pie chart separation only for sex. Then, the second layer could separate age groups and the last layer could separate education level.
2. Triangular representation of the difference between sexes (D) could also be represented as a pyramid — which is a common technique used for census data. In this way, the data for different age groups, instead of education level, could also be incorporated into the overall design created on the emerge phase (7).

— — — — — — — — — — — — — — — — —— — — — — — — — —

3. Group 22 : Novice Village

The blog is well structured and visually appealing. Some things worth to be improved while others should be highlighted. First, let’s start with some positive remarks.

First and foremost, I would like to congratulate the authors for coming up with clear visual designs. The idea to combine a line chart (count of victims versus time) and a map displaying state race/ratio is creative and may bring insights to the reader about possible interactions between number of people killed by the police at a specific location and the underlying demographics of the location.

Second, I like the idea of a circular sketch showing deaths by months. This could unveil seasonality in police behaviours and could give unexpected pieces of information to policy-makers.

This being said, I believe there is room for improvements. I find the sketch in emerge 2 phase, combining line chart with gender histograms particularly clumsy. If you ask me, bar charts should not be mixed with line plots. If you want to combine gender ratio with line chart, I would drop the bars and stick with the lines. Also, instead of creating one separate visualisation by race, you could split your visualisation into N, N being the number of races in your dataset, and create N facets.

Another thing which could be improved is the general definition of your variables. You are referring to “cases” or “ratio of races killed in each state”. However, wouldn’t it be more appropriate to define “cases per inhabitants” or “victims killed in each state, by race” ?

— — — — — — — — — — — — — — — — — — — — — — — — — —

4. Group 2: THEORYOFPATTERNS
Pattern of Intergenerational Supports

A well defined description of design thinking and the process steps are written in the post with the headline “Breaking down our Design Thinking Process”. It is important to give a clear understanding to the readers who are unfamiliar with the concept of visual design.

Intermediate Design Phase

It is seen that the team manage the balance between the diverge and the emerge phase. In the emerge the product of design thoughts are categorized into time and geolocation which look reasonable angles to approach to the Carbone-dioxide and Road Quality measure of interest. It is also interesting that a cross-sectional group as Time-Geo is taken as a distinct category of the designs which may lead interesting advanced visualization ideas. There are provided sketches as the team decided that they are interesting. The first sketch is simple but clear, the second sketch is rather complicated. For instance, the design in which the flower length representing humidity, the color representing the CO2 level and number of petals showing air quality might be hard to follow simultaneously and make an interpretation for a reader who is unfamiliar with the topic. Also, it is hard to find the logical relation between the representative visual component and the variable (exp. petals and air quality). On the other hand the third sketch is clear and easy to make the connection between the representative visual component and the represented variable. Besides, it has indeed a novel design. There is a typo mistake the 4th sketch is shown by “1.” bullet point. The last sketch named “Line and Circle” can be benefited yet it presumes the road quality would change linearly, otherwise the visualization might be hard. At the last section called “Interesting Combinations”, the ideas for the emerge states is provided. It is important to give an idea about the designs of the upcoming post which keeps the reader staying tuned.