Wednesday we're going to meet in the Mac lab in DSH as usual, not at Mitchell Hall. Professor Sanders will give us some extremely useful information about audience and audience analysis.
See you at the Mac lab!
I've been kicking this one around with myself for the last few days, and I think I'm finally at the point where I'm ready to post.
In sum: different cultures handle product advertising (package labels, television adverts, etc.) at different abstraction levels. Arguably, this may reflect the comparative sophistication of the different cultures, or at least their advancement in incorporating abstract - non-touchable - thinking into their daily lives.
Thus, we have the United Kingdom cigarette advertising laws - thou shalt not display, in any tangible way, the name of the product, name of the manufacturer, ingredients, show people using the product, or anything of the sort. Result? The Silk Cut ads: Gold scissors cutting purple silk (the product's colors).
On the flip side, we have Africa in general. With an historically illiterate population, most manufacturers who sell goods in Africa put pictures on packages of what's inside - a very literal one-to-one translation. If the picture is coconuts, you got coconut product inside. If the picture is beans, you got bean stuff inside. And if the picture is a baby, you've just discovered why Gerber's Baby Food doesn't go over well in Africa, and they think the Americans are cannibals.
Continuing with the quotidian use of abstract thinking: the "proof" for the "unintelligent" African used to go like this: when told that all farmers are named Smith and asked if Smith is a farmer, the normal African tribal member would respond, "I don't know; I haven't met him." For any culture that prizes abstract level thinking and intellectual pursuit of hypothetical things - and the sort of culture that would "get" the Silk Cuts word puzzle - this is horrible! However, for any culture whose survival relies on what is tangible and real - and the sort of culture that expects a one-to-one relation between a label and the stuff being labeled - it's perfectly normal, and nothing to be concerned about.
Discussion?
Wow, ten days since the last entry. Things have been very up in the air over the past couple of weeks--unavoidably and understandably so. We have not had many readings and so on to talk about. The graph exercise was enlightening, however. How tough it is to highlight a segment of information from a sea of figures without a clear idea of what numbers will help to display a conclusion that itself is not clear. What were your most difficult issues to conquer on this? I think for our group, trying to merge 3 dimensions of information that each had wildly different units of measure. Comparing % to 1,000s to 100s and trying to draw any kind of a supportable supposition posed quite a challenge.
Tufte’s article was elucidating on the main point: for charts to be truthful and revealing, then the design logic of the chart must reflect the intellectual logic of the analysis. However, the underlying intellectual logic of analysis was not apparent in the engineer’s charts. It appears the engineers did not systematically analyze the data to portray their point of the instability of the o-rings at low temperatures.
The engineers understood the data in a much different sense than the audience. They knew data about test rockets at different temperatures, blow-back, o-ring degradations and dynamics. They knew their information but a coherent representation of their expertise and data was lost in translation to a lay audience. What looks like jumbled data to us, made sense to them from direct observation and analysis. Involved in the details and assuming other people understood their cryptic notes, they forgot to reanalyze their data as Tufte suggests in his text. He is asking the designer (i.e. the engineers) to put themselves in the position of a layperson and understand the logic of presenting their argument. Just as a math teacher would ask you to write down algebra work in solving your problem. In the Challenger case the engineers output observations on o-rings and failed to show the logical connection between their data and intent to cancel the shuttle launch.
Overall, the problem with the engineer’s graph is that they relied too heavily upon specific rockets, o-ring max erosions and data only known to them. In their history of o-ring damage they do not mention temperature which is very relevant to their argument. I was overwhelmed with data and had to rely heavily on Tufte’s explanation as the layperson had to with the engineer’s verbal explanation.
I will agree along with the next person that the visual information presented for the Challenger hearing was confusing and ineffective;however, who are these so called NASA officials to denie the claims of the engineers. Who ever the big wig was that decided to totaly disregard the opinions of the people who actually built the rocket is a freakin idiot. When it comes down to it, a politician made the chief decision in a situation that only the engineers could fully understand. My point being; the actual structure or outline of the information is not as important as understanding where and who it comes from. For example, I could go to a mechanic for an engine inspection. They tell me that my engine is fine but my brakes are bad and if I drive the car I am risking my life. Well they can show me charts and diagrams of how breaks are supposed to work. They can even show me my breaks and how worn they are. That does not matter. Now if I went to Meinike "the break experts" and they said my breaks are fine, I will listen to them. Not the shiesty mechanic. The truth is expertise outways any visual charts or diagrams. Most people rely on the validity of the source, and look to charts and visual evidence for further proof or convincing. But the xurce is the overlying factor in decisions.
My problem with commenting on Tufte is that, in my case, he's preaching to the choir: I don't trust graphs for the most part, or at least, I don't trust graph-makers/chart-makers to really have their audience in mind. Particularly not the New York Times, when they leave New Mexico off their map.
So I'm actually going to wander off and talk about The Visual Display of Quantitative Information, which my mother handed me when I attempted to explain what visual rhetoric was (aiyiyiyiyiyi). It's got one of my favorite charts - the Minard depiction of Napoleon's Moscow advance-and-retreat, tracking army size, temperature on the way back, linking temperature to various points where the army visibly drops in size, and so on. It's beautiful, and the most beautiful thing about it is that once you know three points - Napoleon, Moscow, advance-and-retreat - you can figure out what's going on without speaking a word of French (the entire thing being labeled in French). This is I think wonderful, especially when you compare it to our inability to read the Challenger graphs even /with/ explanatory notes, keys, legends, /and/ extra text.
Although I do wonder what he'd make of the poster I nicked from my mother entitled "A Correlated History of Earth." I shall have to ponder that when I write it up for my image collection.
FWIW -- Kathleen
Getting onto the visual noise thing....
Yes, I agree that the Physics & Astronomy entry page is pretty. But....
I find myself expecting that the Astronomy banner across the top would lead to the Astronomy part of the P&A department. Nope, not a link. It's got great foreground contrast against the darker background, and it's big and central and top and.... it's not a link. Neither is the Physics banner down the left hand side, or the backlit black UNM Physics and Astronomy on the bottom center, or the Videos banner with the arrow that points to ... a button that reads Enter which actually lights up when you run the mouse over it, and is in fact the entry point! So I've now read my way all the way around the entire outside of the fairly spiffy image montage to get to the site entry.
To (possibly mis)quote Douglas Adams: "I can't tell. Anytime I press a black button on the black background a black light lights up black to tell me...." And then of course there's the uber-phantom button in the bottom right that identifies the studio and is a link to that, but since that /isn't/ what I'm logically looking for, being at the P&A site, /that/ makes perfect sense and is really quite spiffy and avoids putting anything I don't need to know out where it could distract me.
But I want the titles that look like buttons to be buttons!
FWIW -- Kathleen
Page 108 is the perfect illustration of the idea it is trying to present. I couldn't read the text with the blaring lined box over it. When I finally covered the box and finished the paragraph I understood: "The noise clouds the flow of information..."
The "shimmer" of statistical graphics does not just draw the eye, it irritates it and causes headaches. I think there are ways to present graphics without causing physical discomfort. I prefer using different shades of color or grayscale. If it is absolutely necessary to use lines or other patterns, I make them bigger.
This article brings home the responsibiltiy a writer or designer can have for providing the correct information. Especially when lives hang in the balance. How often does the wrong information cause an accident, possibly smaller than the Challenger accident, but still just as bad. Mind boggling.
The most interesting part of this Challenger screw up for me was the Congressional hearing. It only reinforced the fact that sometimes the visual isn't even making the point! After ineffective charts and graphs were quite clearly shown to have a serious hand in what happened, Richard Feynman produced this "stunning" visual representation with the ice water! Great presentation, Dick, but that little stunt of yours didn't make the point either!!! So if you had done that for NASA, guess what? It wouldn't have made the point of temperature effects and the whole disaster would have happened anyway! I just found that seriously ironic. Especially since so many people were like, "See? Now THAT was an effective visual display." No, it wasn't!
I actually decided to read the third sequence of Tufte, because part of it involves Dighton Rock, which is only 10 minutes away from the town I grew up in. Oddly enough, I never knew why people made such a big deal about Dighton Rock until I read this-- I thought it was just a park to hike in. Also odd is that I 've taken trips all over the west hiking around looking for pictographs, and all along they were just minutes away from me. Who knew? Anyhow, I really enjoyed the idea of small multiples- how useful they can be was really made clear to me through the pictograph small multiples. Finding all of these figures in the more elaborate renditions of the rock drawings strains the eyes- I found my self getting distracted by the details and the other drawings. But with the one figure extracted, I was able to focus and make more specific comparisons. I think this is perhaps the best example of how small multiples work, because the larger collection of drawings involves images that are large, extensively detailed, and often drawn in very different hands and incorporating different impressions (some details are included in one rendition of the rock & not another). In a sense, it reminds me of the first time I saw one of those cartoon flip-books, where when you flip through the pages fast it looks as though someone is running or swimming or whatever. Looking at those as a kid I was able to see, when I went through each picture individually, the minute variations in each frame that helped compose something larger: the illusion of movement.
Per the items already posted, I thought I'd give a more personal approach to the subject. Tufte, I think, points out what happens when well-meaning and ill-advised scientific thinkers inadvertently convolute data in their graphs, presentations, and papers...
I'd have to agree with Vicky(?) (vqueue), however, that to blame the whole mess on chart junk oversimplifies the data. But the point is well made that the way we present data cannot be taken lightly. Design is meaning, and bad design is misleading. I also agree with the others and with our discussion in class, that while the original charts are a mess, there are still things to be done with Tufte's reworking.
At the Labs, I see a lot of really smart guys make really convoluted graphs. Granted, most of the people in my department have learned the importance of presentation and taken steps throughout their careers to make a good showing. There are those faithful few, however, that help Scott Adams and others to make fun of the socially inept, the woeful chartmakers.
A year or two ago, no Enron employee could have anticipated that all of their personal and business email would be made publically-available by the courts, and yet, it is. The full catalogue of employee email is available via database search. (via NPR)
The all-important issue: Do you as a writer need to consider the fact that ALL of your writing, whether intended for one person or many, might someday be public or might be made public at some time? Can you be comfortable enough with your words to let them go, and ONLY say what you mean, regardless of context? You know what I think What do you think?
I am going to just reiterate what has been said thus far about the Tufte 1 reading. I thought this article was very interesting and provided an inside perspective as to why this tragedy happened. And how it could have been avoided if the charts/graphs would have been clearer and provided more substantial proof that the o-rings would fail in cold temperatures. It is unfortunate that because of someone’s design decision lives were lost onboard the Challenger. I thought the graphs were very unclear and even after reading what they were supposed to indicate and show I was still unable to decipher the information.
As with all our readings so far, it is endlessly difficult to study visuals on pages where they are impossible to see. The Tufte collection is better in that charts in general scan better than more complicated images I've struggled to see. As I said in class today, the early charts were essentially undecipherable to me because I do not understand the material represented on them even though Tufte tells me what it is. The charts he reworks do improve though they are still, in my opinion, pretty crude.
At the same time, he is extremely forceful in presenting his case (chartjunk, smallest effective difference, etc.) that itself uses the example cases as evidence. His purpose, as Thiokol's purpose, is persuasion. His belittling and insulting tone, I think, damages his ethos in this piece. That a poorly designer chart is responsible for Challenger's destruction is a stunning oversimplification of the issues involved in that disaster.
Yet, the principles of design he discusses are logical, sound and useful. The man, in person, must be quite a character. Perhaps too much so, that the personality somewhat overpowers the persuasion.
I just have to say that after finally getting caught up enough to read, and then re-read the Tufte I just had to laugh at the graphics in the essay. Most all of the graphics were absolutely undecipherable for me. There was way to much visual congestion to continue following the reading comfortably. However after finally getting to the reading and being able to then find some time to blog, I feel like I've done nothing more than echo what probably everyone else feels about the reading.
I enjoyed this particular bit of reading as it educated me on the details of the Challenger accident, and I am amazed to find that it was so difficult before and after to express the relevant facts visually. To the degree that I am amazed by this, I suppose it comes somewhat naturally, and so for the first time the importance of training/education for this sort of thing is clear to me, i.e., I see that others--inlcuding the extremely intelligent might not intuitively present the right info to make a case.
I agree that the first revised chart, where Tufte had designated the lauches and pertinent details into columns was easy to understand, but if you can read a graph, I think that the scatterplot expressed the danger even better because it did not rely solely on words and meanings but rather visually displayed the tendency toward damage in cooler temperatures. I do think that the two together would be optimal, though, because they serve complimentary functions.
I found the Tufte 1 reading to be somewhat challenging (no pun intended). It was ALL chartjunk to me, but I suppose that was the point. Given all the confusing statistics and divergent interpretations, it's no wonder the Challenger mission ended in tragedy.
O-rings aside, I took interest in the ear illustration on p. 74, and the accompanying idea of the smallest effective difference. Ban coded pointers! The second illustration (with direct labels) was a clear and effective way to present the image.