> Of Harry Wyeth > Sent: 23 March 2002 08:47 > >I am amazed that NG received only 400 complaints back in the 80s. >With that few they could have simply ignored the whole SI conversion >affair with hardly a dent in their subscription base.
The simple statistics might convince some people that dual metric is a bad policy, they don't convince me. They counted existing customers who left but they did not count the potential customers that failed to join. If I remember rightly, the trade-off is called type 1 versus type 2 error. This is being presented as 'our hands are tied, our readers voted'. I suspect that there are anti-metric people inside the society. They called it an 'experiment'. An 'experiment' sounds like they have not got a good solution or made a decision. It invites a reversal. It all smells like internal committee squabbling and compromise. >After six months no one would care anymore. Yes. You could construct a model for any organisation. Time A: Demand for metric units almost 0% Time B: Demand for metric units almost 100% There is a long transitional period between time A and time B. The loss of business is the cumulative effect of demand and duration. > There is a wonderful NG atlas, circa 1980s sometime, that I own and > treasure. All stats are SI. Interesting, atlases would seem to be dual units now. For some reason the policy for pull-outs is different (dual units) to the policy for articles (US imperial only). Does anybody know why? -- Terry Simpson Human Factors Consultant [EMAIL PROTECTED] www.connected-systems.com Phone: +44 7850 511794
