tag:blogger.com,1999:blog-4093229006254491664.post4863283771141656726..comments2023-08-03T09:55:23.967-04:00Comments on Biased Estimates: Lies or contextually relevant reporting?Tommyhttp://www.blogger.com/profile/14573787040201086607noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-4093229006254491664.post-91085194819324017802014-01-31T09:25:30.167-05:002014-01-31T09:25:30.167-05:00I understand the arguments put forward about not p...I understand the arguments put forward about not plotting misleading graphs. (http://en.wikipedia.org/wiki/Wikipedia:Don't_draw_misleading_graphs) But insisting that zero be on the y-axis is arbitrary. It's context that should dictate the scale of your y-axis (and the range of the x-axis), not an arbitrary value. Consider what zero would mean for the above series, nobody would be employed. That isn't a realistic value that those variables could take. And if I were measure those series as deviations from their means, they'd be centered on zero, but there'd be no new information. Tommyhttps://www.blogger.com/profile/14573787040201086607noreply@blogger.comtag:blogger.com,1999:blog-4093229006254491664.post-21656943646257680072014-01-31T06:03:16.042-05:002014-01-31T06:03:16.042-05:00Using charts to highlight context of any particula...Using charts to highlight context of any particular issue or joie d'affair is similar to a zoom lens on a Cannon EOS; the tighter the shot the more focus, but less perspective. In acute issues where short-term analyses are prudent the tighter lens may be appropriate, but in chronic issue where long-term analyses are required you want the smoothing effect of zero. Budgets, economic activity, social science research, etc., all benefit from the use of zero. Just my two cents...Anonymousnoreply@blogger.com