New answers tagged data-analysis
I've been an instructor for over 18 years and have been doing pre-post test results for over 8 years. An instructor is most likely concerned with improvements between the pre-test and the post-test, and any other demographical data (how many took one or both etc.) would be of secondary interest. Here is a simple chart that I use when I explain the importance ...
"We want to show the improvements in the students who have pre- and post-test data." If that is the ultimate aim of this graph, to show the difference over two tests, then there's no point in graphically showing those who've taken only one part of the test or none of the tests - these can be summarised in a corner/outside the chart with some text (i.e. not ...
The best tip about designing graphs that I can provide is to have a very and simple message that you can see from looking at the visual representation of the data, otherwise it defeats the purpose of clarity in communication and becomes a visual design exercise. Having too much information often means you have to look at other graphs anyway because it is ...
The problem is that you have to fit 3 series of data within one simple chart. I would suggest something along these lines: You display the number of participants as bars, but the success rate is the filling of a bar. Also you may check ds3.js. They have lot's of ideas on visual representation of complex data.
You are correct that star ratings are limited, but they are still useful as a rough guide of quality. On Amazon, I can tell which products I should probably avoid, and which are worth investigating more from the number of stars. On IMDB, I can tell whether a movie is likely to be broadly good or terrible. It's true that I can't gather more fine-grained ...
Why stars in amazon : In amazon it's just a way to assure the user whether the product is god or bad. A star rating on the higher side ( say above 3 with a large number of ratings ) can give a belief to the user that the item has been used by many people and is decent enough to use. One person's 3 and other person's 3 might be different , but still they ...
This type of graph shows min, max, and average values together, in a visually simple way that doesn't require separate lines and labels for each type of value. (Min is zero, I assume in your case.) I don't know the name of this type of graph, but Google seems to think it's a "min-max-average" graph.
I think you shouldn't mix average and maximum information, even in the screen. You said you don't have a mockup of the toggle buttons for avg and max lines on the graph, but I believe that you have something like the following: download bmml source – Wireframes created with Balsamiq Mockups My suggestion is to have 2 different graphs for max ...
I like your initial toggle idea. Can you set it up in a way that it shows as an Average vs. Max toggle for on screen display, and exports as 2 separate side-by-side charts on a PDF? That way you've maximized readability for each display method.
Top 50 recent answers are included