Microsoft Excel

I was confused by the chart-table relationship.
I thought you were asking about style - colours
Then you asked if table & chart match which of course they don't as one is States and the other is People.
What question were you asking?

As for Q18, I can't answer this. The chart has four bars, the table has five rows. There is no match at all.
I chose 'Other'.

IN Q42, I've just noticed that this is the 4th of 5 reports. You've been presenting slightly different colouring schemes for the chart. But the differences have been so subtle I've only just noticed now.  
I think the point of this exercise needs to be much better explained up front.

Just like to say thanks to Microsoft for opening up the communication lines to anyone who whats to makes suggestions, Vista, Office 12, Visual Studio 2005 etc has been a great experiance to see it grow and make input. To me this is a great move to help.

a) help the PR of the products as people understand what is coming and start planning how to implement new technologies
b) gives the product a better "user friendly" touch to it as people express their feelings and frustrations.

The blog site and the newsgroups have been great.

Just done your survey and wish you all the best. Cya round

I agree with the previous comment. The whole test was quite incomprehensible. It seems to be done in the sort of design school where "design" is the opposite of "functionality", as opposed to "a way of improving functionality", as it should be, IMNSHO.

What do I care if the colors "match" if I can't understand the data?

Also, some of the designs were so similar that you're going to get random or semirandom answers. If, on the other hand, they were compared side by side, we could say specifically what we liked and didn't like about each difference.

I really wanted to help out, but my answers unfortunately won't help you. I doubt the rest of the answers will be much more reliable.

Focus of the survey is "How well do the chart and the accompanying table match?" Visual appearance. And the focus is on how well the two objects match, not on which of either looks better or which should be used. Sheesh!

The first example is so simple you don't need the table, just put data labels on the chart. Also the table is transposed wrt the chart, making it twice as hard to see the trends. A more complex table could probably reside in an appendix. The other alternative is to remove the chart and embellish the table with the funky new data bars.

The second table is also transposed wrt its chart, and the legend should be scrapped in favor of a data label on the last point in each series. But you didn't ask how to improve the chart, just how well the table and chart MATCH.

Example 3 very closely resembles example 1, likewise 4 and 2.

And it went downhill from here.

Sysmod and Others,

I put together this survey and I'm happy to answer questions you have about completing it.  As Jon has said, the survey is about how well the Visual Appearance of the table and the chart match.  More specifically, if you were to put a chart with the style (format) and a table with the style in the same document, would you think that the document was visually appealing and professional.  For the survey you should not focus on the data in the table and chart.

One of our goals for the new Theme and Styles feature is to make it easy to create professional, consistent looking workbooks.  We want to ensure that for every chart style that we have included, there is at least one table style that looks good with it in the same workbook.  

For the survey we have chosen some of the styles that we are unsure of and your feedback will help us make final changes to the styles to meet our goal.

Eric,

thanks for your thoughts regarding the survey. I uderstand where you are coming from and what you are trying to achieve, but I don't think you will get the results you need with the current version of the survey.

You say: "For the survey you should not focus on the data in the table and chart." For those of us who make a living presenting figures and facts comprehensivley, there is no such thing is judging charts and tables without looking at the data.

What counts isn't if it looks pretty. What counts is if it makes sense. That's not to say that colors or styles are unimportant. They are very important in helping highlight information. But a chart with 8 items and a table with 20 will not make sense no matter what the colours are.

Alternatively, it may make a lot of sense if the 8 items from the chart have one color in the table and the other 12 have another color. Then I can see how the chart and the table works. That would have been a very interesting setup, and, not coincidentally, an example of how colors help understanding.

When I see charts that look nice but lack meaning my thought is always "Beware! Major bullsh*ters ahead! Stay away!". But maybe I'm in the minority here...

I started the survey but didn't finish it. It became redundant. The obvious problems with the graphics embedded in the article were that they weren't put together with table below chart and the table having a different orientation (time periods down the left side rather than across the top). No point I can see to such a survey if I'm supposed to ignore such glaring faults and just focus on color schemes.

The survey looked like something a child would put together for school. Minimal information content, all touchy-feely. If this is Microsoft's idea of 'professional' . . .

Mere numbers and formulas are confusing and NEED formatting and coordinated color schemes to be made comprehensible. The new Excel! Or should that be New & Improved Excel!

Eric,

I am afraid my responses in the survey will not prove useful to your objective. As an information professional I measure quality in a visual presentation by how clearly, accurately, and succinctly it conveys meaning. High quality presentation imparts insight.

We can already create pretty presentations with existing tools, and with little difficulty. The new stuff is ok but inconsequential. Far more valuable would be tools that help users organize information and discover <i>appropriate representations</i> that communicate effectively. The prevalence of awful charts in publication demonstrates a real need.

To ask about professional appearance without reference to content is to give professionalism a bad name. Look up the word "profess" in a dictionary.
</rant>

I also aborted the survey mid-stream, for much the same reasons as the earlie posters. I don't bleeping care if my charts and tables match. I need to present the data clearly. I would typically not include both a chart and table with the same data in the same report. It's redundant. I don't do color tables, because I don't typically publish in color. When I publish, the table goes into Word anyway and I never ever invest in formatting for publication until after the table is in Word. Charts, on the other hand, I do format for publishing in Excel (they go into word as pictures, because who needs the grief of embedded Excel objects?). But my point is, the whole matching question is trivial. TRIVIAL. But then, I'm one of the 2% of Office users who personalize their menus and toolbars. What the he!! do I know?

All this emphasis on prettification in Office 2007 is giving me ulcers. It makes me feel marginalized.

Much like the other posters, I gave up the survey after about a dozen questions.

If the point of the survey was to measure whether the tables and charts "matched", then you should have asked some schoolkids to complete it.  When you want to know if they've been well-designed and convey the information intelligbly, this is the right audience.

Much like another respondent, my initial thoughts were that a table and chart which (supposedly) show identical information are a waste of space - show one or the other, but show them properly.

One area which would help "match" a table and chart is to use something akin to Conditional Formatting so that the data uses the same colours as the lines on the chart.

Oh, and most of the charts would have been useless if printed on a B&W printer.

I now think I know the meaning of a sentence early in the survey where you said "DO NOT READ THE TEXT".
I took that to mean that you would ask me about the text later, so I did not read the text in the table cells.
But I did, of course, read the data in the chart and table to see how they matched.
Perhaps that is what you meant by "the text"?

Anyway, it's clear that Excel users, data-centric as we are, are not the people to ask a question on design consistency.

I recommend you ask some designers; for example, those who prepare articles in quality magazines that contain tables and
charts and who aim for a consistent visual experience.

The lack of attention to detail in this survey worries me. With such a huge emphasis on making professional "looking" charts, it appears that making "professional" charts has been overlooked.

There is SO much more to charts and tables than color schemes! Professionals who understand the importance of data visualization will continue to do what is necessary to ensure that the data is presented in a clear, concise, readable manner. Hopefully, Excel 2007 doesn't make it even harder than past versions.

My worry is that we will see a flood of eye-candy in reports, presentations, and other documents, without a proportional increase in clarity, conciseness, and readability. Please make sure that the default charts are the most professional, and let the user decide for themselves whether they want to spice things up with alternate color schemes.

The default charts and tables should be printable in black and white (with white backgrounds and black text), easily exported into other programs (not just Office), and appropriate for professional journals, reports, presentations, and other literature. Stay away from gradients due to image size and color limitations of older software.

I don't have anything against nice color schemes, as long as the color is used to enhance data interpretation and doesn't detract from the main message of the table and/or chart.

As of Friday 15:30 PDT, the survey is still up & running. Wonder what the response rate has been? Perhaps the real answer is that it doesn't make sense to ask Excel users about formatting while at the same time asking them to ignore content.

Someday someone at Microsoft may figure out most Excel users really aren't interested in the same things as most PP users.

Most unfortunately, it looks as if the only real thing you are concerned about with this version of Excel is FLUFF.
E.G.,
   * Better shadows
   * Better gradients
   * True 3D effects
   * Glow
   * Soft Edges
   * Reflections
   * Picture recoloring
   * New shapes
How about BETTER CHARTS?
If this is all you're going to do, please give us just one more style option -- <b>a button that will permamently turn off all this stuff and do simple two dimensional flat color charts.</b><p>
And next version please hire someone who knows something about charts -- I suggest Stephen Few, John Walkenbach, Jon Peltier, Andy Pope, Tushar Mehta  -- rather than some fancy graphic artists.

Most unfortunately, it looks as if the only real thing you are concerned about with this version of Excel is FLUFF.
E.G.,
   * Better shadows
   * Better gradients
   * True 3D effects
   * Glow
   * Soft Edges
   * Reflections
   * Picture recoloring
   * New shapes
How about BETTER CHARTS?
If this is all you're going to do, please give us just one more style option -- <b>a button that will permamently turn off all this stuff and do simple two dimensional flat color charts.</b><p>
And next version please hire someone who knows something about charts -- I suggest Stephen Few, John Walkenbach, Jon Peltier, Andy Pope, Tushar Mehta  -- rather than some fancy graphic artists.

I must add my disappointment that the work on chart style seems to be work on chart fluff. How much more welcome a box-and-whisker plot style, or a histogram, or a step line chart, would be. Or perhaps even the information-dense chart styles suggested by Professor Edward Tufte all those years ago in "the Visual Display of Quantitative Information", like the dot-dash plot?

What about a logarithmic range that isn't fixed to a power of ten? Will Excel ever have the capability of reading data labels from ranges, that Lotus 1-2-3 had in the nineteen eighties? I may be out of date, these things may at last have been fixed, but I seem to have watched successive versions leave real gaps in user requirement unfilled, while gadget after gadget has been added to the bells and whistles of "professional looking" charts.

For years Microsoft MVPs have been helping users create approximations of these styles and needs by finessing the limitations of Excel, or creating add-in modules. If only new versions Excel meant that they didn't have to any more.

I've just had a look at some of the previous articles in the series, charting I, II and III, and the irony of all this flash is that the one visual eye-candy feature I think would really be of use to those who are genuinely interested in data presentation (as opposed to glitter presentation), does not seem to be there.

I refer to the prospect of moving beyond the tiny repertoire of line and scatter symbols that Excel has had for more than a decade, the serviceable square, circle, lozenge and triangle, the lonely little cross, the too-thick bar and strange half-bar, the ugly X and even uglier asterisk.

Proper integration of Autoshapes into the data series would be good, by which I mean that they should be formattable from inside the Format Data Series dialogue just as the regular suite of symbols are. They should also behave themselves when imported into other applications like PowerPoint and Word, instead of running amok because they're plotted on a different grid.

I would dearly love a well-behaved triangle that points left, right or down instead of just up, a thinner horizontal line (thinner even than the cross, as hair-thin as the beautiful error bar whiskers that so many of us press into use for purposes they never were meant for), a vertical line to match the horizontal one, and so forth.

And the fluff-lovers could have smiley clown faces, I'm sure they'd like that on their "professional-looking" charts, along with the cones, shadow, lighting effects and 3-D metallic-finish pies.

I've just had a look at some of the previous articles in the series, charting I, II and III, and the irony of all this flash is that the one visual eye-candy feature I think would really be of use to those who are genuinely interested in data presentation (as opposed to glitter presentation), does not seem to be there.

I refer to the prospect of moving beyond the tiny repertoire of line and scatter symbols that Excel has had for more than a decade, the serviceable square, circle, lozenge and triangle, the lonely little cross, the too-thick bar and strange half-bar, the ugly X and even uglier asterisk.

Proper integration of Autoshapes into the data series would be good, by which I mean that they should be formattable from inside the Format Data Series dialogue just as the regular suite of symbols are. They should also behave themselves when imported into other applications like PowerPoint and Word, instead of running amok because they're plotted on a different grid.

I would dearly love a well-behaved triangle that points left, right or down instead of just up, a thinner horizontal line (thinner even than the cross, as hair-thin as the beautiful error bar whiskers that so many of us press into use for purposes they never were meant for), a vertical line to match the horizontal one, and so forth.

And the fluff-lovers could have smiley clown faces, I'm sure they'd like that on their "professional-looking" charts, along with the cones, shadow, lighting effects and 3-D metallic-finish pies.

Derek
EXACTLY!!

Do you think there might be anyone at MSoft we could appeal to?

Anybody got Bill Gates' email?

PingBack from http://ezinefrontpageblog.info/microsoft-excel-give-us-your-thoughts-on-chart-and-table-styles/