TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: Design Resources for Developers?

2 点作者 vital101超过 12 年前
I've been working on a side project for quite some time and have been skirting around the issue of design. I started out using Bootstrap, then purchased a nice looking design from ThemeForest and hacked it to my needs. Even with my lack of design skills, it's actually starting to look pretty nice.<p>The problem I'm having is how to visualize data. Users rate an item on a scale of 1 - 10 for how often that item was required. For instance, a rating of 1 would mean "I needed this item infrequently. Chances are you won't need it either." and a rating of 10 would mean "I needed this item all the time. You will need it.". When the user is entering the rating, it's easy enough to have a quick text description of how it works, however when you're browsing through reviews without that context it's hard to tell what a rating 1 - 10 actually means.<p>Are there any good design resources out there for visualizing data? I need to figure out how to display that kind of data in a way that can be interpreted and understood at a glance.<p>Thanks.

1 comment

vitovito超过 12 年前
There are actually three issues here:<p>1. Are reviewers rating the data using the right scale/format? 1-10 might mean one thing to you and something different to someone else. 1-10 might be inappropriate when you really mean a multiple choice, 1 of N text descriptions. Different types of inputs mean reviewers will <i>think about things</i> differently, and also mean you'll have to process your data differently to get an appropriate visualization.<p>2. What are browsers trying to get out of it? Maybe those ratings, or their descriptions, are helpful in the general case. Maybe they're only really helpful when someone actually wants to make a decision. Maybe browsing <i>by rating</i> instead of by product is how things should be sorted. There's any number of possibilities in how potential browsers might use your service, regardless of your intent.<p>3. Only after you know those things can you decide on an appropriate visualization, and I must warn you that something that looks really attractive and draws people in and gets people to say how awesome your site is is not necessarily the visualization that will be most useful. In addition, most-useful-at-a-glance is not the same as most-useful-after-I-understand-it or most-useful-with-a-lot-of-practice.<p>This sort of thing is why an actual designer is useful. You've got a lot of encoded assumptions and we're trained to break them down.<p>That said, to answer your question, even though it might be the wrong question, there's a great O'Reilly book called "Information Dashboard Design" which explains how to make data visualizations that are actually useful.<p>After you get through that, you might look for an old book called "Using Charts &#38; Graphs" by Jan V. White. It's out of print, so finding a copy might be hard, but it's the opposite of the Tufte books someone will invariably suggest, but which are inappropriate for you. "Using Charts &#38; Graphs" specifically talks about what we now call infographics: telling a story using a chart or a graph.<p>(Tufte's books are inappropriate here because they're textbooks for print people, and while they have lots of good information, implementing their takeaways in electronic form is left as an exercise to the reader. It's way too much work for this situation.)<p>You might also try to dig up the work done by Summize back in 2007/2008, before they were acquired by Twitter.<p>Finally, if you'd like to talk more about this with an actual designer, you might be interested in a design "office hours" experiment I'm running, called UX Hours: <a href="http://uxhours.com/" rel="nofollow">http://uxhours.com/</a>. For comparison, this well-defined question, resulting in advice on things to consider and two places to find out what you need to do, is just about right for a 30 minute consult.
评论 #4801298 未加载