Mike (and anyone else interested)
Thanks. When I modified my calcuations by your suggestion, I got results
pretty close to those published. The data I am using shows the number of
responses we got to each NSS question, by score, 1,2,3,4, or 5. But
these frequencies are rounded to the nearest 5, presumably for data
protection reasons. So, if the Sunday Times were using the true numbers
it would account for the small discrepancies I am still seeing.
There is no discernable discrepancy at institution level, where of
course the numbers are larger, so the rounding to 5 would have a lesser
effect. Even at subject level the discrepancies are less than 1% (I spot
checked 3 subjects). So I think this is the method of calculation, where
'#scores' refers to the sum of responses across all 22 NSS questions:
Numerator:
(#scores'1'x0) + (#scores'2'x1) + (#scores'3'x2) + (#scores'4'x3) +
(#scores'5'x4)
Denominator:
total#scores x 4
Regards,
Jim
Jim Galbraith
Planning Officer
The University of Edinburgh
Old College
South Bridge
Edinburgh
EH8 9YL
tel 0131 650 2097
fax 0131 650 9127
[log in to unmask]
www.planning.ed.ac.uk
-----Original Message-----
From: Academic, financial or space planning in UK universities
[mailto:[log in to unmask]] On Behalf Of MICHAEL MILNE-PICKEN
Sent: 15 September 2006 14:23
To: [log in to unmask]
Subject: Re: Sunday Times University Guide Data 2006
I had a cursory glance at this and tried to work out how they had got a
percentage, but gave up! So I'd like to know too.
I did wonder whether they had subtracted 1 from the denominator and
numerator scores before calculating the percentage; on the grounds you
can't get less than 1.0 and therefore this equates to zero percent
satisfaction,while 5.0 is 100% satisfaction as you cannot get higher.
This would make an average score of 4.0 on NSS, for example, equivalent
to 75% satisfaction [ (4.0-1)/(5-1) = 3/4 ], a score of 3.8 would be 70%
etc. But I then just looked at overall satisfaction and it didn't
match. I didn't think to work out the average of all 22 question
responses.
Have you tried that? Might explain it?
Mike Milne-Picken
Director of Strategic Resource Management and Planning Liverpool Hope
University Hope Park Liverpool
L16 9JD
E-mail: [log in to unmask]
Telephone: 0151 291 3913
Mobile: 0776 4379002
Liverpool Hope University accepts no responsibility for this email, its
contents and any loss or damage arising in any way from the receipt or
use of this email and its attachments
>>> GALBRAITH James <[log in to unmask]> 09/15/06 1:55 pm >>>
Has anybody else tried to replicate the Sunday Times league table
calculation of NSS data into a percentage?
Their published percentage score (as opposed to the weighted league
table points) is "the maximum mark possible if all students had been
happy with every aspect of their higher education". This is based on
answers to all 22 questions in the survey, each of which is of course
answered on a scale of 1-5.
I have tried to replicate this calculation using NSS data we had
already. I took the weighted average response for all questions, all
respondants: (sum of answers ranked '1' x 1) + (sum of answers ranked
'2' x2) etc, the denominator of this calculation being 5 x the sum total
answers to all questions by all respondants. I get a figure
significantly higher than that published, but only have our own data to
test this out with.
Just curious as to whether anyone else has tried...
! Regards,
Jim
-----Original Message-----
From: Academic, financial or space planning in UK universities
[mailto:[log in to unmask]] On Behalf Of David Tucker
Sent: 11 September 2006 16:06
To: [log in to unmask]
Subject: Sunday Times University Guide Data 2006
Does anybody on this list have an electronic version of the Sunday Times
University Guide data in an easily manipulative form such as an Excel
file? Last year we were able to scan in the data and convert it to a
table but this year that is proving problematic, it would appear
somebody would like to make institutional analysis as difficult as
possible! Any help on this would be much appreciated.
Regards,
Dave
D G Tucker
Assistant Planning Officer
University of Surrey
Tel : 01483 683855
Fax: 01483 689549
Email : [log in to unmask]
|