Changing Wants

Research in the Journal of Family Issues by Boxer et al. was illustrated by Max Roser at Our World of Data in the graphic above. From survey results that date back to the 1930s, college students ranked the importance of 18 traits of a prospective mate, via datavizblog, via techinsider.

Here are the ranking results from Boxer et al.:

Their research questions wondered if more recent surveys from 1985, 1996, and 2008 showed increasing importance to men of "good financial prospect." It had only partial support. Their next question wondered if there was increasing importance among women for "good cook and housekeeper" and "desire for home and children". It had no support. Finally, they wondered if there would be decreasing importance among women for "good financial prospect". It had no support.

The graphic above takes their widest view of mate preferences where "mutual attraction - love", "education - intelligence", "sociability", "good lucks", and "good financial prospect" have increased in importance since 1939 for both sexes. "Desire for home, children" has increased for men but decreased for women. And for men "ambition and industriousness" has dropped since 1939. "Chastity" has decreased in importance for both sexes.

X is for .... oh just forget it!

Journalist David Goldenberg of Five Thirty Eight noted the animals most likely used to represent letters in a sample of 50 children's ABC books from 1820 to 2013. He notes that Zebra was used almost exclusively for the letter "Z". But  note the letter "X". So few words begin with "X" that it was most often totally omitted from alphabet books or as Goldenberg says used by authors, "lamely trotting out a fox or an ox and pointing out its last letter." The modern trend seems to be using scientific words such as Xiphias for swordfish.  

Shown these results, one parent mentioned that Xenops, a genus of ovenbirds, was used in at least two of her son's items (books or toys) and was surprised that "D" for dolphin was not higher in the ranking. But I guess it's hard to top Dog and Duck. And Dr. Susess's ABC Book, for the letter D, dreams up a Duck-Dog!

Ordinal Distribution of Letters in Words

We've previously seen the distribution of initial letters in English words. We've also seen the distribution of letter usage in the game of Scrabble. Here is a more complete distribution of letters developed by David Taylor on Prooffreader.com. His distributions are first color coded from pale yellow to deep red indicating the frequency of letter usage from least to most. Next he uses 15 ordinal bins of the relative positions of the letters that begin a word, through the middle of the word, and then the ending letters. Here is his example of how words of varying length were handled.

 

The 4-letter word "four" is apportioned here into only 5 bins. These bin percentages are accumulated across all the words in the Brown corpus via the Natural Language Toolkit. What remains is deciding what aspect of these accumulated percentages of ordinal data to plot to provide an informative display. If the raw percentages are used, comparisons are difficult between frequently used letters like "a" and rarely used ones like "z".

Logs were another possibility he considered, but these add their own interpretation problems.
Normalizing the y-axis so that 100% represents each letter's greatest frequency is another approach. But he argues this makes interpretation difficult since the vertical scales really are not comparable.

And yet another approach is creating an integrated density so that each letter has a density curve with the same area. I think this works best, but he argues that for a letter like "z" with a narrower and taller central density compared to "a" with a broader and lower density, we give more weight to "z", viewing it as having more ink.

In the end he averaged these last two approaches, normalization and integration, to produce his curves. Check out more of his methods for these graphs at prooffreaderplus.