My favorite ways seem to always incorporate a boiling down prior to graphical display, so that the graphical display takes place in a relatively low dimensional space. Top three are factor analysis, culster analysis and correspondence analysis. I find them very helpful as the are quite productive in an exploatory mode. There are brilliant people who can really nail the essential "message of data by looking at conditioned distributions- but you really need to have a sense of what you wish to demostrate as there can be so many different orderings to the conditional distributions. I will try to dig up one of the for an example and I will add it to this note.

If you have not seen Tufte's series on the visual display of data, it is a must. It is really beautiful and very accessible to a broad range of readers.

Maybe this response is not precisely what you are looking for- it is how I like to visually analyze data as opposted to analyze visual data. Still worth it to look at Tufte's books. They are visually awesome!

I always thought there was more room for visual artists to have financially supporting careers focusing on scientific data presentation. The best people who do it now have a strong ability to deconstruct numeric facts and reconstruct them into visual representions. I think developing a four course sequence that makes the analytical side understsndable to visual rtists would really bolster the quality of visual work that goes into scintific articles. I imagine selection of graphical mode of presention (there are many) to be more sensitive to the thematic findings. Woldn't someone with n undertnding of Albers color thory be able to bring out key findings as opposted to some who just likes blue and red bar graphs?

My favorite ways seem to always incorporate a boiling down prior to graphical display, so that the graphical display takes place in a relatively low dimensional space. Top three are factor analysis, culster analysis and correspondence analysis. I find them very helpful as the are quite productive in an exploatory mode. There are brilliant people who can really nail the essential "message of data by looking at conditioned distributions- but you really need to have a sense of what you wish to demostrate as there can be so many different orderings to the conditional distributions. I will try to dig up one of the for an example and I will add it to this note.

If you have not seen Tufte's series on the visual display of data, it is a must. It is really beautiful and very accessible to a broad range of readers.

Maybe this response is not precisely what you are looking for- it is how I like to visually analyze data as opposted to analyze visual data. Still worth it to look at Tufte's books. They are visually awesome!

I always thought there was more room for visual artists to have financially supporting careers focusing on scientific data presentation. The best people who do it now have a strong ability to deconstruct numeric facts and reconstruct them into visual representions. I think developing a four course sequence that makes the analytical side understsndable to visual rtists would really bolster the quality of visual work that goes into scintific articles. I imagine selection of graphical mode of presention (there are many) to be more sensitive to the thematic findings. Woldn't someone with n undertnding of Albers color thory be able to bring out key findings as opposted to some who just likes blue and red bar graphs?

Thanks so much Karen for your thoughtful response and suggestions. Yes, I am looking more along the lines of content analysis of visual media or visual anthropology (e.g., Sarah Pink). I'll check into Tufte's book though!

Hi cohen Miller,

there are different approaches to represent data. Mostly representation depends on type of data[linear, sequential or multi-culture] and it's quantity [Velocity and Variety]. if you will have any sample dataset then i will be happy to suggest you more suitable tools for it.

Regards

Thank you Dr. Sharif for your willingness to respond and examine data.

Gillian Rose's Visual Methodologies; She covers various methods of dealing with visual data. Also look at Jon Prosser 's Image-based Research and Prosser & Loxley:

http://eprints.ncrm.ac.uk/420/

Scott, thanks so much for your suggestions!

I have used the intercultural approach and found it fascinating. Gillian Rose was an inspiration, among others.

Thanks so much Louise for your suggestion about Gillian Rose! This book - http://books.google.com/books/about/Visual_Methodologies.html?id=gnUPNcnYjcIC - looks great. Do you currently use visual methodologies?

Yes I do, using mainly intercultural / historical context approach. If you read French, I wrote a PhD thesis using contextual analysis. I have also published shorter texts. Students and audiences always appreciate such visual approaches.

@Louise I think that's a great point about how audiences appreciate visual approaches. It seems to definitely go over well during presentations as well, at least in my experience.

See my web  site  richardkallweit.com   also all 13 issues of Criss-Cross Art Comminications are now available thru Gildar Gallery  Denver Colorado. We as a group have dealt independently and individually with the issues of pattern recognition ; spatial  analysis and artist mathematician/scientist endeavors . 

While visual approach has always been there lip reading,  body language, signage,

altar pieces, folk tales and so on  it has been until late given second t ier. A person

who has taken a visual approach to scientific is E O Wilson with his discovery of group

evolution. As in ant populations.

An early researcher and compiler of this vision is H.W. Franke.  Leonardo 2000

Kunst im Zeitalter des Computers suhrkamp taschenbuch. page 27 for my work hand done with early copiers for multipliers.