9 posts about research

Populi.net: a mobile phone-based research platform

Apr 11 2008

Now this is what I am talking about: A mobile phone based API for doing things like managing quantitative research projects. Supports a bazillion types of phones. Developed by a South African company. You own your own data. Sounds like a brilliant new project and I look forward to hearing more about it.

The platform was developed for Africa, with all of the barriers that normally come to mind: types of phones used, bandwidth availability and user patterns.Erik Hersman

Now the exciting part is that this isn't just a single application, but rather a platform for application development; that's Populi.net.

Then there is Mobile Researcher(what I think is the first and only app thus far developed for it). Mobile Researcher sounds itself really amazingly cool — you develop surveys on the web and then anyone can take them with their cellphones. They've got an interesting case study up on the web featuring the South African Medical Research Council.

Quoting directly from their case study, it's obvious that there are some exciting ideas going on — I hope to see some further news that reinforces these findings (the only thing that makes me wary is that they have no pricing information, and no screen shots of the survey builder interface):

  • Low cost Nokia 2626 handsets were successfully used by field workers to conduct surveys. Several field workers had never even sent an SMS before.
  • On average, over 400 households were surveyed every day with data available for analysis and reporting before the field workers returned to the field the following day.
  • Research staff and management were able to isolate and rectify issues whilst the study was in progress.
  • Field worker productivity and quality was monitored on a daily basis for training and remuneration purposes.
  • More than 25,000 households were surveyed in under 3 months.
  • In total over 65,000 surveys were conducted.

Via MobileActive.

Kestrel: A Simple Web App for Community Supported Agriculture

Jan 3 2008

I'm just getting started on a new project nicknamed Kestrel.

The basic idea a simple and user-centered web app that helps facilitate ordering, billing and member management for CSA's. Things are JUST getting started and I am soliciting help in doing some feasibility research as well as a basic evaluation of existing CSA management applications.

A CSA, (for Community Supported Agriculture) is a way for the food buying public to create a relationship with a farm and to receive a weekly basket of produce. By making a financial commitment to a farm, people become "members" (or "shareholders," or "subscribers") of the CSA. Local Harvest

So far were in stage zero: Over the holidays I was brainstorming with some of my agri-geek friends in North Carolina, notably tes thraves. (I like to say that tes is to poverty + agriculture issues as Jay-Z is to hip-hop — a badass producer who just makes things happen.) :) So far there's been a lot of excitement about it from both consumers and producers.

  • Stage zero is lots of talk over drinks around the New Year's bonfire, basically. Check.
  • Stage one is research about what real CSA's need.
  • Stage two is getting a few CSA's to pilot test a first iteration for a season.
  • The rest is iterating and improving based on real feedback. This is the hard part. And the fun part.
  • The only real spec so far is an application that is incredibly simple and driven purely by a real understanding of the users' needs.

    There is no timeframe yet. I imagine things could take a year or so; nobody's getting paid by Kestrel.

    Codewise, I've done some simple scaffolding of the application, but really I think the requirements for this type of thing are simple — the codebase is not really the issue. Just a few forms, login/out and billing. So I'm not looking for help from coders as much as I am trying to garner some interest from A) the users of the application, farmers and consumers and B) people with experience in user-centered application design and user testing.

    The goal is a management tool that would simplify the process of ordering food from your CSA, but also serve as an educational model of CSA best practices.

    Right now I'm thinking a hosted solution, almost certainly built in Rails. And of course completely Open and Free.

    The basic use case comes from my mom : she doesn't like very much lettuce in her box. Last year she got six heads of lettuce at a time. So ideally mom could just login and set her preference, pay her bill, update her address, give notice that she's out of town for a month, etc. The farmer then knows exactly how many heads of lettuce to harvest, and can keep the rest in the ground until going to the market on Saturday.

    It's not a new idea, I know. There are several in San Francisco. I haven't seen them yet. But I am sure that they're not as good as they can be and I want to put the users at the front of developing a new open source solution.

    CSA's are great for environmental, social and economic reasons. And they're really just a lot of freaking fun. So if you are a consumer or producer with opinions about what you'd like to see in this type of software, let me know in the comments or unthinkingly-at-gmail.com.

Beautiful Packaging, Ugly Statistics

Dec 16 2005

Recently in my commutes to work I've been using these incredible flash-based, interactive representations of development statistics. (Don't worry, I ride the bus.) These modules, created by a Swedish group called Gapminder, are attractively designed and highly educational. I have long been a great fan of sharp design in the interest of development. These are as great a communication tool as I have ever seen.

Gapminder Screenshot

From the Gapminder website:

"Gapminder is a non-profit venture for development and provision of free software that visualize human development. ... It all started in 1998 from an idea to enhance the understanding of world health. We developed prototype software showing time series of health statistics as moving graphics and varying life conditions as 360¬? photo panoramas from homes, schools and health facilities. From the prototype emerged the Dollar Street project with Save the Children Fund in Sweden and the World Health Chart project with WHO. Within the later project Gapminder developed the free software Trendalyzer that turns boring time series of development statistics into attractive moving graphics. "

ICT4D Africa Scan: An Inventory of ICT Activity in Africa

Nov 23 2005

The ICT4D (Informaiton and Communication Technology for Development) Africa Scan is a serious undertaking that seeks to provide a reference of the major ICT development activities in Africa.

This is a useful place for researchers to begin when attempting to understand the current pace and direction of technology development work on the continent.

People Using a computerThe purpose of this pilot is to experiment with a different way of presenting "who is doing what" in the area of ICT for Development (ICT4D) in Africa. It takes as its starting point a representative sample of funding institutions and maps their ICT4D programs and initiatives in Africa against a backdrop of countries, regions and themes. It also makes an attempt to identify the organisations with which these development institutions partner, and in what areas these partnerships occur. The result is currently not a comprehensive inventory, but an illustration of the potential of using geographic, thematic and partnership information to not only provide an inventory or a snapshot of ICT4D activity in Africa, but to begin to see trends in terms of emerging themes, in terms of countries and regions of specialisation, in terms of partnerships, etc.

Visit: ICT4D Africa Scan

Online Resources for Evaluation Nonprofit Programs

Aug 3 2005

Here's a link to a nice comparison of Google Scholar and Scrius. It points out that Google Scholar has become neglected and is no longer updated regularly. This is a super-unfortunate development; Google is the web's best hope for easy, inexpensive archiving of scholarly research. (In other news, however, Amazon is now offering scholarly articles for a fee. It's easy — they're delivered electronically, but they aren't cheap — about $5.00 per article.)

Here's the link to the Google article: Scholarly Web Searching: Google Scholar and Scirus

Researching Flies, and Colorblindness 101

Jul 13 2005

Drosphilia researchers have a leg up on web designers.

Well, at least they've got a decent explanation of colorblindness. A short paper on colorblind audiences was written a few years ago for researchers presenting their findings on the very latest in the world of flies. The guidelines are easy to understand, and the changes are easy to incorporate.

Here's a bit:

"There are always colorblind people among the audience and readers. There should be more than TEN colorblinds in a room with 250 people. (50% male and 50% female) There is a good chance that the paper you submit may go to colorblind reviewers. Supposing that your paper will be reviewed by three white males (which is not unlikely considering the current population in science), the probability that at least one of them is colorblind is whopping 22%! ... When preparing your presentations (papers, slides, web pages etc.), please take this into account. Here are some comments on how to make figures and presentations colorblind friendly.

I would also recommend a quick look at the Ishihara test for colorblindness for a shot-in-the arm understanding.

And while you're on a colorblindness kick, you'll lose blogger points if you don't read at least Day 12 of the Dive Into Accessibility website, which is the best online introduction to making your website+blog readable by everyone, except illiterates, period.

Ranking Charities = Bad Science

May 4 2005

There is an interesting discussion going on (for some time now) over at the Stanford Social Innovation Review forum about charitable donations and the new "ranking systems" that have emerged to help the public find the most best organizations to give to. The rankings are extremely flawed in the eyes of many, and may be shaping the public valuation of the nonprofit sector in a truly unhealthy, bottom-line-obsessed manner. Here's an excerpt:

"Wouldn't it be nice, as we are sitting down to write our year-end checks to our chosen causes, to have a ratings system to help us make these difficult choices? Indeed, it has long been a dream of many involved with philanthropy and charitable giving to develop such an objective set of criteria to rationalize what is inevitably a highly competitive funding process. Well, several enterprising nonprofit organizations are trying to do just that. The result? Beware of what you wish for."

Read the conversation:Stanford Social Innovation Review: Forum: The Ratings Game

Atlas.Ti: Four Star Research Software

May 21 2004

Do you work with interviews or documents in which you have to make meaningful correlations between themes?

Perhaps you have a lot of data from interviews and you want to know what your respondents associate with a problem.

Atlas.Ti isn't new, but you may have missed the boat. It provides a visual way of organizing and coding data for this type of research projects, especially those involving qualitative data, or data that can't be understood using traditional statistical methods.

All the best research, right?

It's relatively easy to learn, fun to use, and incredibly powerful in the right context.

For a few years now, it has had increasingly powerful support for video and audio coding, which means that you can analyze data from almost any source imaginable. Here's a blurb:

"ATLAS.ti serves as a powerful utility for qualitative analysis, particularly of larger bodies of textual, graphical, audio, and video data. The content or subject matter of these materials is in no way limited to any one particular field of scientific or scholarly investigation. The typical application areas for ATLAS.ti are characterized by a systematic, yet creative approach to analyzing unstructured data." Visit the atlas website.