Showing posts with label Evaluations. Show all posts
Showing posts with label Evaluations. Show all posts

Sunday, February 18, 2007

Online Video Editing Websites

I wasn't able to find a decent video editing freeware for Windows. So I turned to online video editing websites.
I found these two online video editing websites, jumpcut and eyespot. Both pretty do the same things, although I like jumpcut's clip editing controls. Jumpcut has more robust editing features. If you record audio on the video, that's one audio track. Then you can add "background music". With jumpcut, you can fade out the video's audio. So you get a rough version of multitrack audio editing. Both have nice trim functions, and can add special effects. Jumpcut also allows you to remix someone's video. Basically you get access to their project file with timeline and republish it how you want. Imagine a being able to remix a movie with different dialogue or a different ending and then getting to publish it immediately, all in the browser.
Eyespot has a better transitions function. They also seem to have some content from professionals that you can use as well. It's not all user-generated. Definitely check these out. Free uploads, remixing other content, and sharing out to everybody. These are pretty great!
A day after I wrote this O'Reilly Radar reported on an Adobe/Photobucket online video editing site. It's not ready for review. I'll update this post once it's becomes available for use. The main driver in these sites is to create stickiness by editing, an additional feature to draw in YouTube users.

My idea now is to develop an online course using one of these video editing sites to learn how to design instructional videos.

Sunday, February 11, 2007

Conference Presentations

On Paper Presentations
Donata Renfrow and James C. Impara's tips (pdf) provide a very helpful structure for trailblazer studies.
  1. In the Introduction:
    1. Give a teaser, where it fits with prior research;
    2. Build a case, build logic, give background;
    3. Give purpose and importance.
  2. In the Body:
    1. Give enough details on methods and procedures to establish credibility;
    2. Give conclusions with supportive data;
    3. Describe relevant literature, showing how and why your approach is new;
    4. Mention serendipitous findings.
  3. In the Conclusion:
    1. Review study's conclusion;
    2. Note limitations;
    3. Emphasize relevance;
    4. Consider next step.
Gunapala Edirisooriya's analysis of AERA presentations (pdf) provides a few nuggets of insight that I liked. She narrows the talk to the core components:
  1. Background,
  2. Research method,
  3. Data analyses and interpretation,
  4. and Conclusion.
Other tips. Don't waste time on justifying the research method, just explain what you did. Just provide the most important analyses that show the difference or major finding. Be professional, always provide copies of the paper. If you run out, follow up by taking down emails and actually send them as soon as possible. These are people who are interested in your work, cultivate these possible relationships.

Sam Wineburg's 10 tips (pdf) are great for distinguishing between the paper and the presentation. I especially like rules #1 and #8 . Rule #1 is your talk is not your paper. The talk is an advertisement for your paper. Hit the highlights in the 15 minutes you've been given. Tantalize the audience, interest them not bore them. Rule #8 is an academic talk is a diamond. Unlike a newspaper article in which you need to give the reader the whole news in the first section, called an inverted pyramid, you can start small, work your way up to the middle, and wrap it up with a tight summary or conclusion.

The Oceanographic Society's Scientifically Speaking (pdf) is an good overview of tips for both Papers, Posters and even non-scientific audiences. What I really like is their breakdown of visual elements like title and layout of slides. They provide visual examples and non-examples.

My notes on Poster Presentations
So what goes on the poster? For novices, just follow the traditional Science Fair Poster Board three column layout. Major sections from upper left to lower right: Abstract, Research Questions, Methods, Participants, Treatment, Data, Results, and Conclusion. My recommendation is always to emphasize any visual representations of the data, tables, graphs, photographs, concept maps. Too much text is hard to read and process at a poster. You can speak to those. Think about a Movie Poster, how many words do you typically see? Also make sure the font size never drops below 36 point. Keep in mind that someone is standing 3 feet away and probably walking by. Your poster is likely in their peripheral vision.

How do you do it? I always recommend printing your poster on on plotter, 4' x 3'. Hang your poster with double-sided foam tape. Posters are heavy and nothing hampers your ability to explain something that you using one hand to keep your poster from falling down. Get to your session 15 minutes early, skip the session before in order to mentally prepare yourself for your talk. Poster sessions tend to be informal affairs, sometimes food and drinks are served to entice conferees to attend. One time, I was in what was called an interactive poster sessions where there is a 10 to 15 timed talk and then a bell to rotate. This was experimental session that largely failed. The lesson, be prepared for anything. You might have been expected a wall to hang your poster on, and all you've got is a table. Oh, no! Showing up early gives you a chance to figure something out. A poster session sometimes implies that there is a full research paper that supports the work. Sometimes the poster is all that is expected. Find out the presentation requirements from the conference.

References

Online Music Services

finetune.com is a new online music service, that I spent way too much time yesterday playing with. They have stolen the very best features from others into one engaging experience. The way finetune works by the creation of playlists, similar to Apple's iMix. Once you log in, you can create a playlist of 45 songs. It seems like a lot, but they have an I'm Lazy button that will fill out the rest based on the artists already in your list. The website is rocksolid, I noticed only a few bugs. The Flash-based interface of album covers is also iTunes-esque. They have a 30-second preview button for every song, it works much more seemlessly than Amazon's preview feature. Another nice thing about the preview button is that the music is automatically faded or mixed into the current track. This feature, unexpectedly, allows me to compare the music with what's already in my playlist. It helps me to hear if the song will fit in. This feature is as robust as the one that I like at the music kiosks at Barnes & Noble. Better than that, I get to record what I like and hear the full song. I'm still trying to remember a few songs/artists that I heard at B&N that never got written down. The catalog seems pretty full, although a few indie artists and rap groups were either not well represented or there at all. For now, I like it better than AllofMP3 the russian music download store because it's easier to test-drive songs within the context of my other songs in a playlist. I use it for pop songs, indie bands, rappers and world music that generally have a shelf life of a few months for me. The last thing is the social network aspect. While certainly not a unique feature, I found it much better to look at other playlists and similar bands for songs than using the search box, which requires the specific not fuzzy search process. In the end, this is the best experience at me building the top-40 radio station of my choosing. Thanks to Pop Candy's Whitney Mathewson for the heads up.

Pandora is an online music service that provides music recommendations based on artists or songs. Derived from the Music Genome Project, Pandora serves up song based on the "genetic makeup" of the thing that you are searching for. I stopped using the site, a few months ago. The music stations I created tended to throw me off. First, individually they sounded too boring for me. Although I liked the song the station is created on, I kind of like variety, but more than that I just want to hear the song that I like. The Quickmix feature came along, I thought this would solve the problem. It mixed songs from the group of stations I had created. Because I like to listen to music while I work. Being served a song that I hadn't heard forced me to listen, drawing my attention away. Overall, it just didn't work for me. Thanks to Adam Barea for the heads up.

PodShow is different than Finetune and Pandora so it's a bit hard for me to explain. But it is another place to listen to music online. PodShow is really a podcast media company. They work with or run the Podsafe network, where musicians can upload their music. This is the equivalent of the podcasts section of the iTunes store. They have both professionally produced and locally produced podcasts. I have explored too much, but I found a hip hop podcast called D4E Hip Hop Mixtape. The "pj" rummages through the Podsafe network and remixes the songs into a 20 minute "mixtape". There no way I could find such artistry anywhere else. Anyway, I find PodShow's interface very confusing, so I don't really spend a lot of time roaming around. There is just too much for me to sift through. Thanks to Stanford's Entrepreneurial Thought Leaders podcast for interviewing Ron Bloom, co-founder of PodShow, and Ray Lane, venture capitalist.

Tuesday, January 30, 2007

Dreamweaver Alternatives

Aptana is free and pretty robust, however it doesn't have one thing, wysiwyg editing. What I mean is you can't edit the page like a Word document, a very nice feature of DW. You can configure it to view your work in multiple browsers, but all editing is in code. The built-in FTP work very nice, maybe better than DW. The interface is confusing at first. I deleted most of the panels in the workspace. You can add them back as you get more comfortable.


SeaMonkey is an updated version of the old Netscape Suite. It offers wysiwyg editing, but the FTP features are very weak. It's upload only. Although you could use separate software for ftp'ing. It's not as nice a workflow. With every change to a page, means you have to upload a file in a separate program. The interface is easier than Aptana and it seems very stable.

KompoZer is the best of both worlds. It's built on the same codebase as SeaMonkey's web editor and the Nvu program, but it's much better than both. The Ftp features are faster and more stable than in Nvu. The KompoZer development group really just took over for Nvu community. It is for this reason that I would use this software for teaching any sort of basic web development.

An important selection criteria for me in evaluating free and open-source software (FOSS), is to read up on the development community's support of the project. Without their continual efforts, exposed bugs will never get fixed. In my field, I'm often evaluating the viability of software for students to learn in order to get work done. Taking this user-centered approach to product evaluation, I need to make sure that the interface is easy for a beginner to figure out, and that the necessary features are present. Buggy software is just not a good teaching situation, students get frustrated because they have not learned to discriminate errors of their own from errors of the software.