Authoring a PhD Thesis How to Plan, Draft, Write and Finish a Doctoral Dissertation by Patrick Dunleavy

(Brent) #1

more difficult to handle very large numbers (above 1000), and
very small numbers (such as smaller fractions of 1, like 0.0032 or
0.00156). Wherever possible it is best to try and rebase index
numbers to run from 0 to 100, the number range that all read-
ers are most comfortable operating with. In this case, however,
such an effort would mean rebasing to cataract treatments per
1,000,000 people and would have two drawbacks. First, decimal
points would be needed to differentiate some observations from
each other. And second, this measurement unit could be rather
misleading, suggesting to readers that the Scottish health boards
being covered actually havepopulations in the millions, whereas
none of them do. So here rebasing on cataract treatments per
100,000 population delivers tractable numbers running from
217 up to 723. It also makes visible the differences between
observations, but without any clutter of decimal points.


Table 7.2 How Scotland’s health boards compared in
treating cataracts, 1998–9 financial year
Health boards Treatment rates per
100,000 people

Border 723
Tayside 503
Highland 339
Ayrshire and Arran 332
Argyll and Clyde 332
Lothian 318
Greater Glasgow 318
Dumfries and Galloway 317
Western Isles 308
Forth Valley 297
Shetland 282
Grampian 277
Lanarkshire 239
Fife 229
Orkney 217
Mean treatment rate 335
Notes: The range is 506; midspread (dQ) is 55. Two upper outliers, no lower outliers.
Source: National Audit Office, 1999.^4

Upper outlier
Upper outlier

Upper quartile


Lower quartile
Free download pdf